Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Population structure of lake trout (Salvelinus namaycush) In Atlin Lake, British Columbia and contributions… Northrup, Sara 2008

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2008_spring_northrup_sara.pdf [ 2.57MB ]
Metadata
JSON: 24-1.0066241.json
JSON-LD: 24-1.0066241-ld.json
RDF/XML (Pretty): 24-1.0066241-rdf.xml
RDF/JSON: 24-1.0066241-rdf.json
Turtle: 24-1.0066241-turtle.txt
N-Triples: 24-1.0066241-rdf-ntriples.txt
Original Record: 24-1.0066241-source.json
Full Text
24-1.0066241-fulltext.txt
Citation
24-1.0066241.ris

Full Text

        Population structure of lake trout (Salvelinus namaycush) In Atlin Lake, British Columbia and contributions to local fisheries: a microsatellite DNA-based assessment      by Sara Northrup B.Sc. University of Guelph, 1999  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE  in  THE FACULTY OF GRADUATE STUDIES (Zoology)   THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)            February 2008 © Sara Northrup, 2008   ii  Abstract An understanding of the level of both genetic and morphological diversity within a taxon and how that diversity is structured within and across habitats is important when determining the conservation value of that taxon and for successful habitat management programs to be developed.   Atlin Lake is a large lake in northern British Columbia and is one of the largest lakes that contain relatively unperturbed populations of lake trout (Salvelinus namaycush).  As the top aquatic predator, lake trout in Atlin Lake are a key component of the lake’s fish community and are important for local fisheries.   I assayed lake trout from Atlin Lake and other western lake trout populations at eight microsatellite DNA loci and for body morphology to determine: (i) the level of genetic variation present, (ii) the level of substructure that occurs in Atlin Lake, and (iii) whether there was a relationship between the genetic and morphological variation present. STRUCTURE analysis identified five subpopulations within Atlin Lake. Morphological analysis was used to differentiate between the samples collected throughout Atlin Lake.  Cluster analysis of size corrected data separated the fish into two groups making Atlin Lake the smallest lake identified to date to possess more than one morphotype.  Genetic and morphological groupings were found not to be correlated with each other.   Finally, I was interested in whether each of the genetic subpopulations contributed equally to the local fisheries catches.  A mixed stock analysis of samples collected from the commercial fishery and recreational anglers indicated that all of the genetic subpopulations contribute to the fishery along with lake trout subpopulations in the interconnecting Tagish   iii Lake; suggesting that no one subpopulation is being depleted by the fisheries. Continued genetic monitoring, however, is necessary to see if the trends in fishery contribution are temporally stable.  Future studies should focus on understanding the source of the morphological variation and maintenance of genetic substructure.     iv Table of Contents  Abstract................................................................................................................. ii List of Tables ........................................................................................................ v List of Figures ..................................................................................................... vii Acknowledgements……………………………………………………..………………ix Chapter 1: General Introduction............................................................................1 Importance of Genetic Diversity ........................................................................1 Lake Trout Divergence through Glaciation ........................................................4 Current Knowledge of Lake Trout Variation ......................................................5 Lake Trout Conservation ...................................................................................7 Atlin Lake, British Columbia ..............................................................................9 Atlin Lake Community Biomonitoring Program................................................10 Thesis Objectives ............................................................................................11 Chapter 2: Population Structure of Salvelinus namaycush in Atlin Lake, British Columbia.............................................................................................................13 Introduction .....................................................................................................13 Materials & Methods........................................................................................19 Results ............................................................................................................24 Discussion.......................................................................................................32 Chapter 3: Phenotypic Diversity as it Compares to Genetic Diversity in Salvelinus namaycush in Atlin Lake, British Columbia…………………………………………45 Introduction .....................................................................................................45 Materials & Methods........................................................................................49 Results ............................................................................................................51 Discussion.......................................................................................................53 Chapter 4:  Contribution of Salvelinus namaycush Populations in Atlin Lake to Local Fisheries....................................................................................................61 Introduction .....................................................................................................61 Materials & Methods........................................................................................64 Results ............................................................................................................66 Discussion.......................................................................................................67 Chapter 5: General Discussion ...........................................................................72 Population Structure of Atlin Lake Lake Trout .................................................72 Phenotypic vs Genetic Diversity ......................................................................72 Fisheries Management....................................................................................74 Conservation ...................................................................................................76 References .........................................................................................................78   v  List of Tables  Table 1: Sample site names, number of samples, locations, watersheds, elevations and surface area................................................................................91  Table 2: Microsatellite locus names, references, annealing temperatures, and size range of alleles in base pairs for western lake trout used in the current study. ............................................................................................................................92  Table 3: Summary of allelic variation at eight microsatellite loci in lake trout. Number of alleles per locus (A), expected heterozygosity (He), observed heterozygosity (Ho), allelic richness (Ar ), and number of individuals genotyped (N) for each loci per population.  Sample locations are defined in Table 1. ........93  Table 4: Estimates of pairwise FST values (mean over 8 loci) for 19 western lake trout lakes. Population codes are defined in Table 1. .........................................96  Table 5: Log likelihood scores from STRUCTURE for all sample lakes. The boldface values represents the highest likelihood...............................................97  Table 6: Analysis of molecular variance within and among western lake trout populations. ........................................................................................................98  Table 7:  Estimates of effective population size of all sample lakes....................99  Table 8: Log likelihood scores from STRUCTURE for Atlin Lake.  The boldface value represents the highest likelihood. ............................................................100  Table 9: Summary of allelic variation at eight microsatellite loci in lake trout of each subpopulation within Atlin Lake.  Number of alleles per locus (A), expected heterozygosity (He), observed heterozygosity (Ho), allelic richness(Ar), and number of individuals genotyped (N) for each loci per population. ...................101  Table 10: Estimates of pairwise FST values (mean over 8 loci) for five lake trout subpopulations within Atlin Lake. Boldface values are significantly greater than 0. ..........................................................................................................................102  Table 11: Estimates of effective population sizes of Atlin and Tagish lake subpopulations..................................................................................................103  Table 12: Contingency test results of geographic units within Atlin Lake and Atlin genetic subpopulations .....................................................................................104  Table 13: The mean pairwise identities for relatedness for Atlin and Tagish Lake subpopulations (A-E and A-D, respectively). ....................................................105   vi Table 14: Summary of allelic variation at eight microsatellite loci in lake trout of each subpopulation within Tagish Lake.  Number of alleles per locus (A), expected heterozygosity (He), observed heterozygosity (Ho), allelic richness(Ar) and number of individuals genotyped (N) for each loci per population..............106  Table 15: Estimates of pairwise FST values (mean over 8 loci) for four lake trout subpopulations within Tagish Lake. Boldface values are significantly greater than 0........................................................................................................................107  Table 16: Log likelihood scores from STRUCTURE for Tagish Lake. The boldface value represents the highest likelihood. .............................................108  Table 17: Contingency test results of spawning locations within Tagish Lake and Tagish genetic subpopulations .........................................................................109  Table 18:  Log likelihood scores from STRUCTURE for Atlin and Tagish lakes. The boldface value represents the highest likelihood. ......................................110  Table 19: Loadings of eigenvectors on four principal components for the residuals of 36 body measurements on 104 lake trout from Atlin Lake.............111  Table 20: Contingency test results assessing an association between geographic units and morphological groups. ....................................................112  Table 21: Contingency test results assessing an association between genetic subpopulations and morphological groups........................................................113  Table 22: Genetic mixture analysis results of 101 commercial samples and 33 recreational angling samples of lake trout  within Atlin Lake. Mixture values represent the estimates contribution of each subpopulation within Atlin and Tagish lakes. The averages and standard deviations are based on simulations from 5000 replicates. The boldface values represent the subpopulations with the greatest estimated contributions. ......................................................................114  Table 23: Genetic mixture analysis results of 101 commercial samples and 33 recreational angling samples of lake trout within Atlin Lake. Mixture values represent the estimates contribution of each subpopulation within Atlin and Tagish lakes. The averages and standard deviations are based on simulations from 5000 replicates.The boldface values represent the subpopulations with the greatest estimated contributions. The mixture analysis was conducted using rigid reference populations with admixed genomes removed (> 0.5 admixed) for both Atlin and Tagish subpopulations. ......................................................................115       vii List of Figures  Figure 1: Location of lake trout sample sites examined in this study.  Population codes are given in Table 1................................................................................116  Figure 2: Allele frequencies in lake trout from four major watersheds and assayed at eight microsatellite loci..................................................................................120  Figure 3: Principal components analysis of allele frequency variation at eight microsatellite loci among all sample lakes.  The numbers correspond to the population codes listed in Table 1.....................................................................121  Figure 4: Isolation-by-distance analyses for lake trout populations in western Canada. ............................................................................................................122  Figure 5: Allele frequencies in lake trout sampled from Atlin Lake and assayed at eight microsatellite loci.  A-E refer to the five subpopulations found in Atlin Lake and correspond to those mentioned in Table 9.................................................126  Figure 6: Map of the distribution of lake trout subpopulations within Atlin Lake according to STRUCTURE assignments.  The ----- indicates the separation between geographical units.  The value to the right is the sample size for each section of the lake. ............................................................................................127  Figure 7: Allele frequencies in lake trout sampled from Tagish Lake and assayed at eight microsatellite loci. TA-TD refer to the four subpopulations found in Tagish Lake and correspond those mentioned in Table 13. .........................................131  Figure 8: Principal components analysis of allele frequency variation at eight microsatellite loci among subpopulations of lake trout in Atlin and Tagish Lake (1-5 corresponds with Atlin subpopulations A-E and 6-9 with Tagish subpopulations A-D). ........................................................................................132  Figure 9: Map of the distribution of lake trout subpopulations at spawning beds within Tagish Lake according to STRUCTURE assignments.  The value to the right is the sample size for each section of the lake..........................................133  Figure 10: Known or Suspected Spawning Locations in Atlin Lake and Tagish Lake (Atlin Community Fisheries Working Group 2001) ...................................134  Figure 11:  Locations of colouration measurements used in morphological analysis.  Each square represents an area 40x40psi........................................135  Figure 12: Landmarks used to compare shapes of lake trout sampled within Atlin Lake: 1 -  anterior tip of snout, 2 - posterior tip of maxilla, 3 – center of eye, 4 – top of cranium at midpoint of eye, 5 – posterior of neurocranium above tip of   viii opercle, 6 – anterior insertion of dorsal fin, 7 – posterior insertion of dorsal fin, 8 – anterior insertion of adipose fin, 9 – dorsal insertion of caudal fin, 10 – midpoint of hypural plate, 11 – ventral insertion of caudal fin, 12 – posterior insertion of anal fin, 13 – anterior insertion of anal fin, 14 – insertion point of pelvic fin, 15 – insertion point of pectoral fin, 16 – connection between gill covers...................136  Figure 13: Head depth (HD), mid body depth (MBD), caudal peduncle depth (CD), and body length (FL) measurements of lake trout from Atlin Lake. .........137  Figure 14:  Transitions in landmark position contributing to PC1 of morphological variation in lake trout.  Each arrow is an exaggeration of the displacement in the X and Y coordinates of each landmark position. ...............................................138  Figure 15: Models generated by McCLUST to determine the most probable number of clusters (components) within the lake trout morphological data based on Bayesian Information Criterion (BIC).   Each line and symbol represents a different model (see McCLUST for model definitions).......................................139  Figure 16:  Atlin Lake lake trout samples plotted along the first and second principle components (PC1 and P2).  Ellipses encircle about 95% of the measurements present in each cluster of individuals as classified in the MCLUST analysis.  Triangles correspond to morphotype 1 and squares to morphotype 2. ..........................................................................................................................140 Figure 17: Map of the distribution of the two morphotypes among four geographic localities  within Atlin Lake.   The value to the right is the sample size for each section of the lake. ............................................................................................141  Figure 18:  Results of DFA indicating the level of overlap between the two morphotypes. ....................................................................................................142  Figure 19:  ANOVA results of colouration comparisons to (a) genetic subpopulations (b) morphological clusters and (c) geographic unit.  Map indicates the geographic units within Atlin Lake.   Pictures indicate the extreme colouration differences found within Atlin Lake....................................................................143  Figure 20: Map of the commercial fishery (squares) and common angling fishing (circles) locations within Atlin Lake ...................................................................144   ix Acknowledgements  I would like to thank Taku River Tlingit Fisheries Department and the Yukon Territory Government for funding my thesis work.  I would like to thank my supervisor Eric Taylor for challenging me and sharing his knowledge of evolutionary genetics and native fishes.  I thank Sean Rogers for his patience in demonstrating the use of morphological data analysis and his long discussions on phenotypic variation.  For assistance in the lab, I would like to thank Patrick Tamkee and Jennifer Gow who, along with Lesley Harris, were very helpful in fish biology discussions.  I thank Dolph Schluter and Luke Harmon for their assistance in learning analysis programs.  I would like to thank Mark Connor and Susan Thompson for helping to organize my fieldwork and for gathering samples from local anglers in the Atlin area.  For assistance in the field I would like to thank the Yukon Territory Fisheries Group; Aaron Foos, Angela Milani, Nathan Ferguson, as well as, the Tlingit Fisheries Group; Richard Erkhart, and Jerry Jack.  I would further like to thank all the people who went to the trouble of collecting and sending me samples: Great Northern Fish Company, Tlingit Fisheries Department, Yukon Territory Fisheries Department, Paul Sparling, Ray Pillipow, Lesley Harris, Jonathan Witt, Gary Hill, James Williams, and Randy Kelcher.   I would also like to thank my fellow graduate students in the Taylor lab for being patient sounding boards and for making for a fun as well as productive working environment.  Finally, I would like to thank my family and friends for encouraging me to continue my studies and for tolerating the endless lake trout conversations.   1 Chapter 1: General Introduction   Importance of Genetic Diversity The origin, extent and patterns of genetic diversity within a species are questions evolutionary biologists have focused on for years (Frankham et al. 2002).  Why does it matter whether genetic diversity is present?  Further, why is it important if this presence of diversity is known? Genetic diversity can affect growth, survival, fertility, developmental rate and the ability of those individuals to develop properly (Allendorf and Leary 1988), adaptation to changing environments, and their ability to make use of a variety of habitats.    Given that levels of genetic diversity can influence the persistence of populations (Frankham et al. 2002) there is a general need to study genetic diversity of contemporary populations in order to predict possible changes to population viability in response to changes in genetic variation owing to environmental fluctuations.  One aspect of genetic variation is population subdivision, the partitioning of a species into two or more independent or semi-independent genetic subpopulations that may exist in the same or different habitats. To manage a population effectively, knowledge of the amount of genetic subdivision within a single habitat or location is required.  Without knowledge of the population structure, management could be designed for general needs of the overall population or species rather than considering each of the potential subgroups as a separate entity, which may respond independently to environmental variation and harvest and management practices (Ryman 1983; Wenburg et al. 1998).   Species as a whole are more successfully   2 maintained, as are the diverse habitats across which they reside, once the level of subdivision is known (Frankham 1995; Avise and Hamrick 1996). Why use genetics to assess diversity?  Genetic characteristics tend to be more discriminating and accurate when detecting population structure than previous methods employed, such as phenotypic differentiation (Li et al. 2002).  For instance, microsatellites, a class of highly variable codominant DNA loci, have been used to assess levels of population subdivision from across a range of plant and animal taxa (Raybould et al. 1999; Neilsen et al. 1994; Paetkau and Strobeck 1994; Forbes et al. 1995; Morris et al. 1996; Wenburg et al. 1998; Taylor et al. 2003).  The use of microsatellites often allows greater resolving power, repeatability, and accuracy in the detection of intraspecific population structure (Bruford and Wayne 1993; Wenburg et al. 1998) and has become a standard assay in many aspects of conservation genetics (Sunnucks 2000). My study employed microsatellites to assay the variation in lake trout (Salmonidae: Salvelinus namaycush) to determine the level of subdivision within Atlin Lake, BC.  Char – Biology Lake trout, Salvelinus namaycush, is a member of the Salmonidae family (Pisces) which includes salmon, trout, char, grayling and whitefish.  Lake trout, however, are not true trout, but rather are char and included in the genus Salvelinus which includes six major species; Salvelinus alpinus, S. malma, S. confluentus, S. leucomaenis, S. namaycush, and S. fontinalis (Westrich et al.   3 2002).   Modern day Pacific salmon and char are thought to have existed for at least the last 6 million years (Power 2002).  The current distribution of char includes areas that have been influenced, in terms of both the climate and topography, by the Pleistocene glaciations over the last 1.8 million years.  These repeated glaciations have provided multiple opportunities for the alteration, isolation, and reconnection of peripheral freshwater habitats across the Holarctic and for their fish faunas in different areas to have diverged from one another (Hewitt 1996; Magnan et al. 2002; Westrich et al. 2002).   Hybridization between char species has been noted between S .alpinus and S. fontinalis (Bernatchez et al. 1995), S. alpinus and S. namaycush (Wilson and Hebert 1993) and S. malma and S. confluentus (Baxter et al. 1997).  Due to secondary contact between divergent lineages of char and subsequent hybridization the phylogenetic relationships between taxa of the genus can be complicated and at times controversial (Westrich et al. 2002).  The history of char, as well as the well- documented morphological, ecological and genetic variability within the genus, makes these species excellent models for studying evolutionary processes (Magnan et al. 2002). Lake trout is endemic to North America north of about 40oN latitude and go by a variety of common names: lake char, lakers, tongue trout, mountain trout, mackinaw trout, and most commonly, lake trout (Scott and Crossman 1973). Lake trout is predominantly a piscivore, and reside in the deep, cold water of northern Canada, Alaska, the Great Lakes and parts of New England (Gunn et al. 2003).  Its most obvious identifying feature is its large size (average is about   4 5kg, 50cm in length, maximum recorded 46.3kg, 126cm in length) and deeply forked caudal fin (Scott and Crossman 1973).  Lake trout are slow growing and late maturing (6-8 years) and display extensive morphological variation.  They vary not only in size and body shape, being “lean” or quite “broad”, but also in their colouration, which ranges from silver to brown to green (Scott and Crossman 1973).  These different morphotypes are usually associated with differences in habitat and behaviour (Brown et al. 1981).  Lake Trout Divergence through Glaciation Lake trout dispersal has been limited by the effects of the Pleistocene glaciations. At least five different glacial refugia are thought to have contributed to the current populations, suggesting that vicariance provides one explanation for the mitochondrial DNA variation found among populations (Wilson and Hebert 1998).  The exact locations of lake trout refugia are difficult to determine due to the changes in the climate following the end of the ice age (Wilson and Bernatchez 1998). The movements of the Laurentide and Cordilleran ice sheets during the Wisconsonian glaciation (75,000 to 10,000 years ago) have most recently shaped contemporary lake trout habitats (Lindsey and McPhail 1986). The glaciers caused deeply scoured lake basins, along with a series of temporary drainage systems, which acted as dispersal routes (Gunn et al. 2003). The glaciers retreated earlier in western Canada than in eastern Canada, with the resulting melt water facilitating lake trout dispersal along these glacier- scoured routes, which connected the proglacial lakes inland. The melting ice also   5 reduced salinity levels in along the Arctic coast allowing lake trout to tolerate and disperse in these waters (Lindsey and McPhail 1986).  Secondary contact between refugial groups has been suggested as another source for intrapopulation diversity owing to the geographic locations of diverse populations and that of former proglacial lakes (Wilson and Hebert 1998). Repeated glaciations have caused the reduction of habitats and separated populations for long periods promoting divergence between the populations (Wilson and Hebert 1998; Wilson and Bernatchez 1998).    In addition, some genetic diversity within populations may have been lost due to sustained bottlenecks during glacial intervals, which may also promote divergence among populations (Wilson and Bernatchez 1998). When the glaciers retreated, however, refugial groups came into contact with each other which also could have contributed to the intrapopulation diversity currently observed (Wilson and Hebert 1998).   The major mitochondrial lineages that have been resolved have largely allopatric distributions, suggesting dispersal from separate refugia, i.e., from the “Bering”, “Nahanni”, Mississippi” and “Atlantic” refugia (Wilson and Hebert 1998; Wilson and Bernatchez 1998).  Current Knowledge of Lake Trout Variation Of the lake trout populations studied, the largest amount of phenotypic variation has been seen in the Laurentian Great Lakes (Krueger and Ihssen 1995). There are three morphotypes reported from the Laurentian Great Lakes; “lean”, “humper” and “siscowet” morphotypes are recognized based on facial   6 characteristics, body fat content and spawning time (Burnham-Curtis and Smith 1994).  Apparently, Lake Superior is the only remaining lake that has sympatric populations of each of the three morphotypes (Krueger and Ihssen 1995). The first documented case of sympatric populations of lake trout outside the Laurentian Great Lakes was in Great Bear Lake, NWT.  It contains two populations, one of which is similar to that of the lean form found in the Laurentian Great Lakes; however, these two populations are not spatially separated like those in the Laurentian Great Lakes (Alfonso 2004).  Initial evidence suggested that there is a genetic component to this differentiation and the presence of intermediate phenotypes in the Laurentian Great Lakes suggests the gene flow occurs among the populations (Burnham-Curtis and Smith 1994). Further studies identified substantial genetic differentiation among sympatric and allopatric populations in morphology, early developmental rate, physiology, and in allozyme and DNA-based allele frequencies (Eshenroder and Krueger 2002). Currently only a few native populations remain in Lake Huron and Lake Superior. In the mid 20th century, a drastic reduction in lake trout population sizes and diversity occurred in the Great Lakes due to overfishing, invasive sea lamprey (Petromyzon marinus) predation, and habitat destruction (Guinand et al. 2003). As a result of the attempt to restore the diversity of lake trout in the Great Lakes many genetic studies have been conducted to understand the levels and origins of diversity. Microsatellite DNA analyses of young-of-the-year lake trout have provided robust estimates of population-specific reproductive success in restoration programs in the Great Lakes (Page et al. 2003).   7 Gene flow hinders genetic differentiation from occurring among populations (Avise et al. 1986).  The prevention or lack of gene flow, whether due to physical barriers or behavioural reproductive isolation, provides a possibility for genetic divergence to occur (Slatkin 1985). Individual lake trout have been reported by Ihssen et al. (1988) to disperse up to 300 km from spawning beds to surrounding feeding areas. Despite their potential for long distance dispersal, biochemical genetic analysis and tagging of mature lake trout in large lakes suggest that they exhibit precise homing to spawning grounds, which can maintain genetic isolation between populations (Ihssen et al. 1988).     In addition, population size has been correlated with the level of genetic diversity in that larger populations typically maintain greater genetic diversity particularly at neutral loci (Nei et al. 1975; Ihssen et al. 1988).  Furthermore, the number of individuals that colonized a lake after the last glaciation could have a profound effect on the degree of contemporary diversity (Ihssen et al. 1988).  By contrast, relationships have also been found between genetic diversity and habitat types, as well as, community complexity (Soule 1973; Ihssen et al. 1988).  Lake Trout Conservation Lake trout are important members of aquatic communities across their range. First, lake trout are dominant predators of north temperate freshwater lakes in the Nearctic (G. Low, Department of Fisheries and Oceans, Hay River, NWT personal communication). Whether lake trout are sought in commercial recreational, or subsistence fisheries depends on the area.   For instance, in   8 western Canada lake trout are one of the most desired of freshwater fishes in Nunavut, Northwest Territories,  Yukon, and British Columbia (BC).  By contrast, Alberta has limited fisheries, primarily due to the general lack of large, cold-water lakes that are the preferred habitat of lake trout. The degree of exploitation of lake trout lakes varies across the range as well.  Declines in abundance and loss of populations are occurring in a number of areas leading to the loss of morphological and genetic diversity (Guinand et al. 2003). In addition, invasions by non-native species and human activities can place populations at risk (Frankham et al. 2002) and have caused problems for persistence of lake trout in particular (Guinand et al. 2003).   Populations with a greater level of genetic and phenotypic variability may be especially important to conservation because their higher degree of genetic variation may be important for adapting to changing conditions (Frankham et al. 2002).  The study of genetic diversity can provide early signs of populations at risk, such as potentially “inbred populations” (Lynch 1997), as well as alerting managers to any particular genetically unique populations that may exist (Hedrick 2001).  Clearly, an understanding of lake trout genetics is applicable to range-wide conservation initiatives (e.g., Committee on the Status of Endangered Wildlife in Canada, COSEWIC, status reviews) cannot proceed without knowledge of populations across a wider portion of its range than has been investigated at present.   9 Atlin Lake, British Columbia Until recently lake trout have primarily been studied in depauperate (e.g., Great Bear Lake) or disturbed lakes (Laurentian Great Lakes).  In order for depressed populations to be successfully restored knowledge about the level of genetic and morphological variation from a self-sustaining, relatively unperturbed natural system would be useful.  Ranges in the levels of natural diversity and effective population sizes of healthy populations vary from species to species, therefore, a baseline for lake trout is needed in order to help guide restoration in perturbed lakes (Gunn et al. 2003).  In BC, lake trout are considered to be an unthreatened species (T. Down, BC Ministry of Environment, Victoria, BC, personal communication), but there have been few quantitative studies to support this supposition.   The relatively pristine environment and healthy lake trout population(s) of Atlin Lake, therefore, make an excellent case study for the examination of lake trout biodiversity. Atlin Lake is an irregular shaped lake spanning an area of 775sq km in northwestern BC extending into the Yukon Territory. It is the largest natural lake ecosystem in BC and possesses diverse aquatic resources, but very little is known about the system.  During the Pleistocene glaciations, this area was buried under ice and it has been suggested fish from multiple refugia repopulated it post-glacially (Wilson and Hebert 1998).   In fact, three glaciers remain at the southern portion of the lake and supply the lake with melt-water and turbidity, in the form of glacial flow, in the areas of glacial influence.  A portion of Atlin Lake is contained within the boundaries of the Atlin Provincial Park that ensures its   10 relatively unperturbed state.   In an area known for its hunting and recreational fishing Atlin Lake has been left unperturbed by overfishing and other anthropogenic disturbances and is presently zoned as “natural environment” (Atlin Community Fisheries Working Group 2001). Lake trout in Atlin Lake are an important component of the freshwater fisheries resources in this area and are used for recreational, subsistence, and limited commercial fishing in the northern, Yukon portion of the lake as well as being the focus of an important local recreational fishery.  The first study of the lake was a limnological survey performed by Withler (1956 cited in Atlin Community Fisheries Working Group 2001) in response to hydroelectric development that was planned on the lake.  Studies on Atlin Lake’s fish community thus far consist of presence/absence information (Lindsay et al. 1981), limnological characterizations (Kirkland and Gray 1986; Shortreed and Stockner 1983) and, most recently, lake trout abundance, levels of exploitation, and spawning information (Atlin Community Fisheries Working Group 2001).   It has been noted that there are a variety of phenotypes within the lake (Mark Connor, Tlingit First Nation Fisheries Department, Atlin BC, personal communication), but how this variability may be organized by habitat, spawning area, and associated with spatial and temporal genetic variability is completely unknown.  Atlin Lake Community Biomonitoring Program My study contributes to a larger project aimed at improving management decision-making in Atlin Lake specifically regarding the fisheries. The Atlin Lake   11 Community Biomonitoring Program is run by the Taku River Tlingit First Nation Fisheries Department and is seeking to give a biophysical profile of the lake. Other studies in this program include mapping and monitoring of lake trout spawning habitats, an angler survey, contaminant analysis, otolith collection and aging, as well as compiling the existing information including traditional knowledge about lake trout.  Knowledge and maintenance of diversity is of vital importance if we want to prevent exploitation and loss of genetic diversity in lake trout.  Genetic diversity can be used as a measure of long-term population health; establishing a baseline of genetic diversity can aid in determining the health of the population(s) and those potentially requiring conservation attention (Frankham et al. 2002). Diversity should also be conserved for its inherent value as reflecting the evolutionary legacy of a species (Bowen 1999).  Thesis Objectives My thesis had three primary goals. First, I wanted to provide a survey of both microsatellite DNA and phenotypic variation (morphology and colouration) in lake trout from Atlin Lake.  These data were used to determine the level of population subdivision that exists in lake trout within Atlin Lake.  Second, I wanted to test if there is any association between the different phenotypes observed and genetic differentiation.  Finally, I wanted to use my genetic data to quantitatively assess whether each population in Atlin Lake contributed equally to the various fisheries. These data would contribute to a greater understanding of the structure of diversity in Atlin Lake S. namaycush and provide some guidelines for more   12 effective management and conservation. In contrast to the vast majority of previous research on lake trout, which has been conducted on highly degraded and impacted populations of the Laurentian Great Lakes, my study is the first to provide a detailed analysis of genetic diversity in western lake trout from an unperturbed ecosystem.  My study contributes to a broader understanding of diversity across the range of this species of temperate lakes of North America.   13 Chapter 2: Population structure of Salvelinus namaycush in Atlin Lake, British Columbia Introduction Intraspecific genetic differentiation has become of great interest to biologists and conservationists over the past several decades.  The presence of genetic diversity is of great importance because the loss of genetic diversity is thought to increase the risk of population extinction and reduce evolutionary potential (Frankham 1995). Genetically depauperate populations are more common in nature than was originally thought (Caughley 1994). Declines in abundance and loss of populations are occurring in a number of species leading to the loss of morphological and genetic diversity (Guinand et al. 2003).   There has been a large-scale loss of diversity and abundance along with disruption of population and community structure in a number of fish due to environmental change and human interference (Rahel 2000). Most species are subdivided to varying degrees into genetically distinct subpopulations and a major component of conservation is the identification of distinct intraspecific groups, which require preservation. These distinctions are fundamental for successful conservation decisions; lack of accuracy of such information, has been one of the foremost reasons for failure of some conservation programs (Ryman and Utter 1987; Schwartz et al. 2007).   14 There have been several definitions of “population” over the years and many definitions depend on what specific aspects one is considering. For example, many of the evolutionary definitions concern potential for intermating whereas the ecological definitions tend to focus on habitat (Waples and Gaggiotti 2006). For my research, I base the definition of populations on of the level of genetic variation within and between samples and the degree of interbreeding between them (Frankham et al. 2002).   In other words, a population is a collection of individuals found in a particular area that have a greater probability of mating with other individuals of the same species in the same area, relative to individuals from other localities, such that genetic differences (from isolation) accumulate between individuals from different areas. Management practices vary depending on how populations are defined.  Whether to manage a species as one or several populations can depend critically on the definition of what constitutes a population (Ihssen et al. 1988). One of the principal characteristics of a population that determines the level of genetic diversity, and of programs to sustain it, is the effective population size (Ne); that is the number of individuals in an ideal population, which experiences the same rate of change as the natural population (Schwartz et al. 1998).  Ne not only determines the rate of loss of neutral genetic variation, but also the rate of fixation for both deleterious and favourable alleles and the rate of increase of inbreeding within a population (Frankham et al. 2002). Consequently, Ne estimates are critical for effective conservation strategies (Schwartz et al. 1998). Ne can be determined in some species based on field data, but this can take   15 considerable time and be quite difficult with wild populations’ especially aquatic species (Primmer 2005).   Although some species can manage with fairly low Ne’s others cannot, therefore, studies and regular monitoring allow the distinction between normal population size fluctuations and ones that require conservation intervention (Primmer 2005).  An example of this is the lower effective population sizes often found in species from freshwater environments compared to those of marine species suggesting a greater susceptibility to extinction of the former species and, hence, perhaps they might warrant more conservation attention (Primmer 2005).  Declines in effective population size  of some species in the Laurentian Great Lakes have been suggested to result in extinction of subpopulations (Guinand et al. 2003) making the identification of any substructure and the quantification of the Ne a high priority for currently healthy populations. Sources of genetic distinctiveness are many and vary from species to species and population to population.  It is difficult to determine the degree of genetic divergence that any one morphological or ecological difference may represent (Phelps and Knudsen 1987). For instance, sympatric and allopatric lake trout populations have shown genetic differentiation in morphology, early development rate, physiology, allozyme and DNA based allele frequencies (Ihssen et al. 1988). A significant degree of genetic isolation between populations has been observed, as have geographic patterns in allele frequencies (Ihssen et al. 1988). Identification of genetically distinct populations based on morphology is often misleading. In many cases, it can be impossible due to the environmental   16 influences on morphology, which could generate morphological similarity in genetically distinct populations and vice versa.   Tagging programs have been used to study the independence of populations and the levels of interlocality dispersal, but such methods are not always practical or even possible (Ryman and Utter 1987).    Various biochemical and molecular genetic assays have risen to the forefront to address questions of population subdivision owing to their relative ease of use and unambiguous genetic control (Pella and Milner 1987; Amos and Balmford 2001). Recently many fish species, including many salmonids, have been the focus of conservation studies that use population distinctiveness to separate native from introduced fish and to identify distinct populations (Gunn et al. 2003).  My study focuses on the population(s) of Atlin Lake.  Atlin Lake is a natural lake system used for commercial and recreational fishing and has not been subject to   fish stocking.   Due to its large size and diverse nature, both in terms of habitat and lake trout phenotype, it was suspected to contain multiple populations. Questions raised by such a possibility include: what is the nature and origin of each population’s distinctiveness?  Is it due to isolation prior to colonization of the lake (i.e. multiple sources of colonization) or due to local adaptation to their present habitat? Lake trout populations have been found in Atlin Lake since they were colonized postglacially from the Nahanni and Berinigian refugia approximately 10,000 years ago (Wilson and Hebert 1998).  Many subdivisions found in fish species are due to old divergences resulting from long periods of isolation caused by glaciation   17 and further differentiated by postglacial isolation and adaptations to their present environment (Power 2002).  The many episodes of isolation while in separate refugia followed by secondary contact between the divergent lineages promoted tremendous morphological, ecological and genetic variability within the species making lake trout a good model for studying the evolutionary process (Magnan et al. 2002).  In addition to the importance of lake trout from Atlin Lake to the local community this particular population(s) is of enormous interest to the greater study and conservation of the species because few populations in the western part of the geographic range of the species have been studied, especially for pristine areas like Atlin Lake. In rainbow trout (Oncorhynchus mykiss) and brook trout (Salvelinus fontinalis) genetic diversity has been positively correlated with lake size (Griffiths 1997).  Sympatric morphotypes of lake trout have to date only been found in one lake (Lake Mistassini, Québec) outside of the Mackenzie and Laurentian Great Lakes which raises the question; how large must a lake be to contain multiple populations?   Perhaps the most valuable information lake trout from Atlin Lake can provide is an answer to the question: what level of diversity may be present in a natural system?  It has often been stated that loss of diversity increases the risk of population extinction (Frankham 1995); however, without a baseline for comparison how are we to determine if any particular lake’s population is at risk? Objective In this chapter, I examine genetic variation at eight microsatellite loci to study the level of diversity in western Canadian lake trout. Many early studies on lake trout   18 in the Laurentian Great Lakes employed protein allozymes and were able to successfully differentiate between hatchery and native populations (Grew et al. 1994).  Allozyme surveys (Dehring et al. 1981; Ihssen et al. 1988) showed evidence of structuring within and between morphotypes found in the Laurentian Great Lakes.   Mitochondrial DNA was also employed for this purpose and found to identify refugia of origin with great success (Vitic and Strobeck 1996; Wilson and Hebert 1998) but was not as accurate when using small sample sizes to examine fine scale population structure (Piller et al. 2005).  Angers and Bernatchez (1996) illustrated that microsatellites were highly polymorphic in lake trout and may be more informative and have greater discriminatory power than both allozymes and mtDNA for population structure analysis. Microsatellite markers are highly polymorphic thus exhibit considerable intraspecific variation to quantify genetic structure in char (e.g., Angers and Bernatchez 1998; Costello et al. 2003) and are especially useful for resolving structure in populations that have only recently diverged (David and Capy 1988).  Microsatellites are further used because they are suspected to be neutral loci and therefore allelic variation probably plays no part in differential survival.  Historical populations of lake trout have been assayed with the use of microsatellites to reveal that populations differed depending on their basin of origin (Guinand et al. 2003).  Additional studies (Page et al. 2003) have confirmed that microsatellites are an excellent diagnostic tool for differentiating between lake trout populations and are sensitive enough to distinguish between sympatric lake subpopulations.   19 In particular, I looked at the population structure found in Atlin and Tagish lakes. I collected data on genetic variation among lake trout populations in BC, NWT and Yukon to address the following hypotheses:  First, that Atlin Lake lake trout originated by postglacial recolonization from more than one refugium. Secondly, that due to this historical repopulation and present day resources in the lake, there is a high degree of genetic diversity within lake trout found in Atlin Lake as compared to other western lake trout populations.  Finally, due to the unique features of the lake and the high degree of genetic diversity I hypothesized that population substructure has developed or in the process of developing within Atlin Lake.  Materials & Methods Sample Collections Lake trout tissue samples were collected from 2000 until 2006 from 19 lake populations throughout BC, the Yukon and Northwest Territories (Table 1, Figure 1).  Lake trout were caught using small mesh gillnets deployed at 1km intervals for one hour.  Samples were weighed, measured and the adipose fin tissue was removed and stored in 95% ethanol.  In 2005 and 2006, samples from the Atlin Lake survey were photographed for morphological and colouration analyses (see Chapter 3).     20 Microsatellites Total genomic DNA was extracted using a QIAgen DNeasy kit.  DNA was precipitated and resuspended in Tris-EDTA (pH 8.0) buffer then stored at -20oC. Microsatellite loci were selected from previously published primers found to be polymorphic in salmonids.  Microsatellite loci were screened and chosen based on clarity of amplification and visible variability.  Eight microsatellites were selected (Table 2), fluorescently labeled and used for all analyses.  Polymerase chain reactions (PCR) were performed in 10ul volumes in MJ PTC 100 thermal- cyclers.  The PCR mixtures consisted of approximately 100ng of DNA, 0.4mM dNTPs, 5uM reverse primer, 2uM fluorescently-labeled forward primer, 1.5mM KCl (NEB combined with 10x Buffer) and 0.5 units of Taq polymerase (NEB). The PCR program was: 1 cycle  (95°C/ 3 min), 5 cycles (95°C /30 sec, TA+2oC/ 45 sec, 72°C /45sec), 35 cycles  (92°C / 45 sec, TA/ 45 sec, 72°C / 45sec), and 1 cycle (72°C / 5min), where TA is the annealing temperature(s). The microsatellites were multiplexed and assayed on a Beckman-Coulter CEQ 8000 automated capillary sequencer/genotyper. DNA Analysis Hardy-Weinberg Equilibrium (HWE) and linkage disequilibrium were tested using GENEPOP version 3.1 (Raymond and Rousset 1995).  Hardy-Weinberg equilibrium for each locus-population combination and linkage disequilibrium for locus pairs within each population were tested using Fisher’s exact test in which p-values were estimated using the Markov chain method.  An initial p-value level of significance was α = 0.05 and corrected for multiple comparisons using   21 Bonferroni adjustments (Rice 1989).  Basic descriptive statistics of sample size (N), number of alleles (A), observed heterozygosity (Ho), and expected heterozygosity (HE) were performed using TFPGA version 3.2 (Miller 1997) while allelic richness (Ar) was determined using FSTAT version 2.9 (Goudet 2001). GENEPOP was employed further to test for population differentiation for all pairs of populations over all loci combined using log-likelihood (G)-based exact tests (Goudet et al. 1996). Weir and Cockerham’s (1984) Fst (Ө) values were calculated using ARLEQUIN version 3.0 (Excoffier et al. 2005) to test population differentiation via pairwise comparisons.  Microsatellites typically mutate following the stepwise mutation model (Ohta and Kimura 1973) and although Rst was developed as an analogue of the statistic Fst to follow the stepwise mutation model (Slatkin 1995), Fst has been found to outperform Rst in recently diverged populations when relatively low sample sizes and a small number of loci are used (Gaggiotti et al. 1999), therefore, focus will be placed on Fst-based statistics. With the use of the program PCAgen (Goudet 1999) a principal components analysis (PCA) was conducted on allele frequency data to compare genetic differentiation among populations.  The variation across all eight loci in all populations is summarized and oriented with major axes of variation, accounting for independent measures of genetic differentiation. Microsatellite variation was partitioned into its components by performing an analysis of molecular variance (AMOVA) as explained by Excoffier et al. (1992) and employing the program using ARLEQUIN version 3.0 (Excoffier et al. 2005).   22 I tested different hierarchical arrangements of samples such as watershed groupings, PCA groupings, as well as, Atlin Lake subpopulations compared to Tagish Lake subpopulations. Population structure The number of populations in (a) Atlin Lake (b) Tagish Lake were estimated by employing the program  STRUCTURE version 2.1 (Pritchard et al. 2000).  This program conducts an analysis of population divergence assuming no specific a prior structure.  A Bayesian clustering algorithm, Markov chain Monte Carlo based approach assumes a model of K populations, characterized by allele frequencies at each locus and assigns individuals probabilistically to each of these populations.  This analysis then uses a likelihood approach to determine the most probable number of K populations using a general genetic inheritance model to minimize Hardy-Weinberg and linkage disequilibrium within the cluster groups.  Through multiple trials and following Evanno et al. (2006) suggestions I employed a ‘burn-in’ period of 10,000 iterations and further replicated with 10,000 iterations; longer runs within increased replication did not alter the results. Twenty simulations of each K value ranging from one to eight using admixture and correlated frequencies models were run to calculate the log likelihood of data. I also followed the suggested protocol of Evanno et al. (2005) to determine the ad hoc statistic ∆K, which was used to estimate of the true number of populations.  Overestimates of K have been noted when using the point where the log likelihood plateaus to determine the most likely value for K, especially as incremental changes made interpretation difficult (Pritchard et al. 2001).  The ∆K   23 statistics was developed to alleviate this difficulty and is based on the rate of change in the log likelihood between the preceding and successive K values (Evanno et al. 2005).  If K = 1 has a log (K) higher than the other potential K values, then the sample is considered to contain only one cluster. By contrast, when the most likely value is not K = 1, then the K with the highest ∆K is the most likely number of clusters.  This was used to determine the number of subpopulations found within my lakes of interest and these subpopulations were used for the remaining analyses. Effective population (Ne) size for each of the lakes and subpopulations was estimated by measuring the association of alleles at different loci. NeEstimator (Peel et al. 2004) calculated linkage disequilibrium and the correlation among alleles at different loci.  The advantage of calculating Ne from linkage disequilibrium rather than the more common temporal method is that only one sample period is required (Bartley et al. 1992). Isolation-by-Distance The correlation between geographic distance and genetic distance (Fst) and its significance was tested to determine if the observed genetic structure could be explained by the isolation-by-distance model (IBD: Slatkin 1993).  TFPGA was employed to perform the Mantel test (Mantel 1967) and significance of any correlation was determined using the default setting of 999 matrix permutations. Google Earth (http://earth.google.com/download-earth.html) was employed to calculate the distances between locations.   24 Lake trout samples from Atlin Lake were collected during a lake survey; therefore, the lake was subdivided into four large geographic units to analyze subpopulation and geographic correlation.   Contingency tests were performed using the program PAST (Hammer et al. 2001) to determine (1) if the genetic subpopulations within Atlin Lake were associated with the four geographic units established and (2) if the genetic subpopulations within Tagish Lake were associated with distinct spawning locations. Individual pairwise relatedness indices of the Atlin and Tagish lake subpopulations identified by STRUCTURE were calculated with IDENTIX (Belkhir et al. 2002).  IDENTIX has a number of different estimators for determining the relatedness of individuals one being the identity index (Mathieu et al. 1990). Identity defines relatedness as a measure of the expected proportion of loci that would be homozygous in offspring from the selected pair of individual. I calculated individual relatedness values since the sub-populations resolved by STRUCTURE did not resolve marked geographic patterns; however, because my sampling was not based on actual spawning locations, examining individual relatedness values may resolve finer patterns of genetic differentiation.  Results Intralocality Variation Microsatellite variation across 818 individuals at eight loci was assayed.  The number of alleles ranged from one (Sco102) to thirty (Smm22) across all populations with an average of 14.8 alleles per locus for the entire study region   25 (Table 3) and an average of 5.3 alleles per locus per lake.  Observed heterozygosity ranged from 0.17 to 0.91 with an average of 0.50 across all loci and populations (Table 3).  The two most variable loci were found to be Smm22 and Ssa197 with 30 and 24 alleles globally, respectively.  Only 14 out of 152 (8 loci x 19 populations) tests showed statistically significant deviations from HWE and all were heterozygote deficiencies. Seven such deficiencies were from Atlin Lake (4 loci) and Tagish Lake (3 loci); the remaining samples were found to be largely in Hardy-Weinberg equilibrium.  Tests for linkage disequilibrium resulted in none of the 532 tests being significant and in disequilibrium.   Consequently, all loci are considered to represent independent measures of genetic variation and divergence. The level of variation within populations varied greatly. Expected heterozygosity, over all eight loci, ranged from a high of 0.75 (Atlin Lake) to a low of 0.30 (Bednesti Lake) (Table 3).  Fourteen populations were found to have at least one of the eight loci fixed with Sco102 being the most commonly fixed locus.  No more than two loci, however, were fixed in any one population. Between watersheds, there were no fixed allele differences across the eight microsatellites studied; however, there were several alleles found to be unique to individual watersheds (Figure 2 A-H).  Unique alleles were found predominantly in the Yukon and Fraser rivers’ watersheds.   There was a strong, positive correlation between lake area and allelic richness (r = 0.68 P = 0.0012).     26 Population Structure There were 171 pairwise comparisons of FST summed across all eight loci between populations for differences in allele frequencies and all but 34 of these were significant at α = 0.05 (Table 4). The average FST across all 19 populations was found to be 0.212; ranging from 0.144 to 0.284.  Most of the non-significant differences involved Muncho Lake were the sample size was very small (N = 6). A principal components analysis (PCA) of allele frequency for all lakes revealed two groupings (Figure 3) which was later confirmed in STRUCTURE (Table 5). The Fraser River watershed lakes and Traviallant Lake (NWT) grouped together with the remaining samples comprising the second group in both tests. Lake trout were found to display a pattern of isolation-by-distance (IBD) within watersheds, but not across multiple watersheds.  All 19 populations were initially included and a test for IBD was found not significant (Figure 4A).  The populations were then divided according to the two PCA clusters and tested for IBD, which was also found to be not significant (Figure 4C & D).  When, however, the Fraser River watershed was analyzed individually it was found to follow an IBD pattern (r = 0.47, P = 0.025; Figure 4B). Genetic variation was further examined by grouping the various populations by watershed excluding Muncho Lake due to it being the sole member of its watershed and its small sample size.  AMOVA indicated that variation among the watershed groups (19.58%, P < 0.001) exceeded that among populations within watersheds (13.76%, P < 0.001, Table 6).  Further tests were conducted based on the groupings suggested by the principal components analysis (Figure 3).   27 Here, there was a narrowing of the difference among PCA groups (3.3%, P < 0.001) as well as that among populations within PCA groups (6.0%, P < 0.001). The effective population size (Ne) within the sampled lakes ranged from 21.1 to several thousand (Table 7).  Not all of these lakes, however, were sampled thoroughly while others had additional subdivision within the lake resulting in potentially skewed estimates, which should be viewed in only a relative sense. Effective population size was significantly and positively correlated with lake size (r = 0.66, P = 0.002). Atlin Lake One hundred and eighty two individuals from Atlin Lake were genotyped at eight microsatellite loci.  STRUCTURE analysis indicated the presence of five sub- populations (A-E) within Atlin Lake (Table 9), with many of the individuals possessing some alleles from multiple subpopulations indicating either gene flow and/or recent structuring. Evanno et al. (2005) and Latch et al. (2006) used simulated data to test the robustness of ∆K statistic and the ability of STRUCTURE to correctly identify the number of subpopulations.  By employing the ∆K statistic along with the STRUCTURE data the correct number of subpopulations was identified in all cases when FST was 0.03 or higher (i.e., within the range in Atlin and Tagish lakes).  These simulations, along with the common use of this statistic in population structure analysis (Gow et al. 2007; Zamudio and Wieczorek 2007) provide confidence in my estimate of at least 5 subpopulations within Atlin Lake. There is, however, no current method to statistically assess the differences between different log likelihood values in the   28 present context, i.e., that 5 subpopulations are the true number of subpopulations within Atlin Lake. Rather, K = 5 is only the most likely given the present data. With this caveat in mind, the remaining analyses were conducted using K = 5; however, the presence of some substructure is of greater importance in these analyses than the actual number of subpopulations. The number of alleles ranged from two (Sco102) to fifteen (Smm22) across all sub-populations with the average of 11.6 alleles per locus (Table 9).  Observed heterozygosity ranged from 0.19 to 0.97 with an average of 0.64 across all loci and populations (Table 9).  The most variable loci were found to be Smm22 and Ssa197 with 26 and 18 alleles globally, respectively.  Only one of the 40 (8 loci x 5 populations) tests showed statistically significant HWE deficiency; all subpopulations were found to be in HWE.  Tests for linkage disequilibrium resulted in none of the 140 tests to be significant and in disequilibrium. There were ten pairwise comparisons between Atlin Lake STRUCTURE- identified subpopulations all of which were significant (Table 10) with an overall global FST of 0.046.  The effective population size of Atlin Lake was calculated initially prior to determining the level of substructure within the lake.  Once subpopulations were identified the underlying assumptions used to calculate Ne (ie. no substructure) were violated and, therefore, each individual subpopulation’s Ne was recalculated and ranged from 29.1 to 807.3 (Table 11). Each Atlin Lake subpopulation did contain at least one unique allele, but none were fixed (Figure 5 A-H).  Subpopulation C possessed the most unique alleles at five (Smm22 *234 base pairs, Ssa197 *275, Sco107 *232, Sco215 *311 *319),   29 subpopulation A followed with three (Smm22 *148 *152, Ssa197 *287), subpopulations D and E both possessed two unique alleles (Ssa197 *219, Sco107 *256) (Ssa197 *199, Sco215 *291), respectively, and subpopulation B contained one unique allele (Ssa197 * 271).   Of the two unique alleles that subpopulation D possessed, one allele (Sco107 *256) was found to be unique not only within Atlin Lake but among all the lakes studied. Contingency tests (Table 12) identified a significant (P = 0.0035) geographic partitioning of the genetic subpopulations A-E within Atlin Lake; however, as can be seen in Figure 6 there is still a fair amount of mixture throughout the lake. Subpopulation A is found mostly in the north, subpopulation B in the south, while subpopulation E is the more central and subpopulation D is the evenly distributed throughout the lake.     A summary of the individual relatedness indices (Table 13) showed that subpopulation D is the most distinct from all others with B, A, E, and C progressively less distinct.   The two most closely related subpopulations were D and B with A and E being the most distinct from each other. Tagish Lake Microsatellite variation across 89 individuals at eight loci was assayed.  The number of alleles ranged from one (Sco102) to sixteen (Smm22) across all sub- populations with the average of ten alleles per locus (Table 14).  Observed heterozygosity ranged from 0.09 to 0.97 with the average of 0.67 across all loci and populations (Table 14).  The most variable loci were found to be Smm22 and Ssa197 with 23 and 19 alleles globally, respectively.  Two of the 32 (8 loci x 4 populations) tests showed statistically significant HWE deficiencies; the   30 remaining were found to be in Hardy-Weinberg equilibrium.  Tests for linkage disequilibrium resulted in only one significant deviation out of 112 comparisons. There were six FST pairwise comparisons of allele frequency differences between Tagish sub-populations and all were significant (Table 15).  Effective population sizes in Tagish Lake ranged from 94.3 to several thousand (Table 11). Substructure was found within Tagish Lake with four subpopulations resolved to be the most likely by STRUCTURE and the use of the ∆K statistic (Table 16).  A contingency test (Table 17, Figure 9) showed a significant relationship between the subpopulations and the spawning beds.  The remaining samples, those not collected from the spawning beds, with known collection locations were too few in number to be representative of the lake therefore the degree of geographic differentiation is unknown.   Individual relatedness (Table 13) indicates that subpopulation TAG-B was the most distinct from all others  TAG-A and D were found to be the quite distinct from each other while TAG-C and D were the most closely-related populations. The subpopulations within Tagish Lake were each found to possess unique alleles, however, there were no fixed allelic differences among the four subpopulations (Figure 7 A-H).  Subpopulation B contained three unique alleles (Smm22 *164 *238, Sco107 *252) while subpopulations A and D contained two unique alleles (A – Sco102 *168, Ssa197 *223; D – Ssa197 *227 *287) and the remaining subpopulation C contained one (C – Smm22 *160). Between Atlin and Tagish lakes there were found to be no fixed allele differences; however, Atlin Lake did possess ten alleles not found in Tagish Lake   31 (Smm22 *144 *148 *152 *168 *234, Sco19 *176, Sco107 *256, Sco215 *291 *307 *319) while Tagish Lake contained nine unique alleles (Smm22 *239 *238, Sco102 *168, Ssa197 *283, Sco107 *248 *252, Sco2 *167 *169 *177 *187). FST values were found to be higher in Tagish Lake (Table 15) than in Atlin Lake (Table 10). Nine subpopulations were determined to exist across both lakes through a STRUCTURE analysis (Table 18) which also confirmed that Atlin and Tagish lakes’ subpopulations were distinct from each other. A principal components analysis between the Atlin and Tagish lakes’ subpopulations showed Atlin Lake subpopulations C, D and E were tightly clustered with Tagish Lake subpopulations C and D, however, neither axis was found to be significant (Figure 8). An analysis of molecular variation was performed grouping all Atlin Lake populations and all Tagish Lake populations into two groups.  The AMOVA indicated that only a small portion of the variation found was attributable to differences between lakes (0.42%, P = 0.086), but slightly more was attributable to differences among subpopulations within lakes (3.58%, P < 0.001), and most variation was attributable to differences among individual fish within subpopulations (95.99%, P < 0.001, Table 6).   32 Discussion The ability to detect genetic substructure within populations has greatly increased over the last decade and assisted in populations being managed more effectively.   Genetic assays have been used to identify quickly and accurately the level of genetic diversity, population fragmentation, gene flow, effective populations size, and accumulation of deleterious mutations, each of which provides information for developing a management scheme appropriate to each species and habitat (Frankham et al. 2002).  This has been seen in the selection of hatchery strains for restoration efforts in the Laurentian Great Lakes (Krueger et al. 1989). In this study, I sought to determine the level of subdivision of lake trout within Atlin Lake as well as to compare the level of genetic diversity and allelic richness of western lake trout populations to those of the more heavily studied and perturbed systems of eastern Canada.  These data will be used as baseline levels of genetic diversity for more effective conservation and management decisions for lake trout both in Atlin Lake and more generally. Western Lake Trout Vs Eastern Lake Trout Lake trout are considered to have lower heterozygosity when compared to other salmonid species (Dehring et al. 1981; Vitic and Strobeck 1996).   My results, however, found that western lake trout had comparable variation, and even higher than in one char species, (e.g., coastal bull trout, S. confluentus HE = 0.35, Taylor and Costello 2006), but are much lower when compared to others (e.g., brook char, S. fontinalis HE ~ 0.70, (Angers et al. 1996), Arctic char, S. alpinus , HE = 0.82,  (Bernatchez 2002) and Atlantic salmon, Salmo salar, HE ~ 0.76,   33 (Neilsen et al. 1999)). Heterozgosity is dependent on the loci selected, however, and many of these studies used a few of the same loci and all were averaged over many loci allowing for comparisons to be made with greater confidence. Genetic diversity at microsatellite loci appears, however, to be slightly higher in western lake trout populations (Table 3) relative to that of eastern lake trout. Average expected heterozygosities over all my study lakes was 0.51±0.056 and was consistent with those reported from contemporary Laurentian Great Lakes’ populations of 0.51±0.036 (Guinand et al. 2003).    An average number of 5.3 alleles per locus per lake was found to be slightly higher compared to that reported in eastern lake trout of 4.5 alleles per locus (Page et al. 2003) which may be due to the population declines known to have occurred in the east.  It is possible that the proximity of my samples to multiple glacial refugia could be a reason for a slightly higher level of diversity since the further a population is located from the putative refugial source of repopulation, the greater is the potential for losses of variation during recolonization from founder effects and bottlenecks (e.g., Power 2002; Costello et al. 2003).   Alternatively, the generally reduced fishing pressure on the western lakes than those in the east may have allowed larger contemporary populations sizes and greater retention of genetic diversity (B. Landry, US Geological Survey, Great Lakes Science Center, Oswego, NY, personal communication). An average FST was found to be 0.212 over all 19 study lakes while a level of subdivision (GST) of 0.041 was found among contemporary populations in Lake Huron and Lake Superior (Guinand et al. 2003).   Although GST and FST cannot   34 be compared directly it is evident that western lake trout show a much higher degree of subdivision.   It has been noted, however, that FST tends to be lower when continuously-distributed populations (e.g., lakes Huron and Superior) are compared relative to more widely-separated or completely isolated populations (e.g., my study lakes, DeWoody and Avise 1999). For instance, average microsatellite-based FST in salmonids, across different degrees of physical isolation has been found to be 0.27 (Hendry and Stearns 2004).  Secondly, FST increases as lakes become further apart (Costello et al. 2003; Taylor et al. 2003) therefore it is not surprising that my study lakes would be more differentiated as they span from the Northwest Territiories to central BC. Within Western Lake Trout The extent of microsatellite variability was not uniform across all 19 western lake trout populations sampled.  When the data were pooled across loci, lakes were found to be highly divergent from one another while the subpopulations within Atlin and Tagish lakes displayed only slight differentiation from one another. Analyses based on mtDNA have indicated that Atlin Lake and Tagish Lake were most likely colonized postglacially by fish from the Nahanni and Bering refugia while the Mackenzie River watershed (e.g. Traviallant Lake) was colonized by fish from the Mississippi refugium, and the Fraser River watershed was colonized by fish from the Bering refugium (Wilson and Hebert 1998).   My analyses of allelic diversity and population structure analysis supports the idea that Atlin and Tagish lakes were repopulated by fish from multiple refugia.  For instance, Atlin and Tagish lakes’ had the highest degree of allelic richness of all the lakes   35 studied and had much higher levels of heterozygosity than found in historical (0.466±0.011) and contemporary (0.513±0.036) populations in the Laurentian Great Lakes (Guinand et al. 2003). Lake trout from Atlin and Tagish lakes were found to be very similar to one another, whereas, the Fraser River watershed lakes and Traviallant Lake were found to form a cluster separate from Atlin and Tagish lakes suggesting that they were repopulated by fish from separate refugia and differed from those in Atlin and Tagish lakes. My multilocus microsatellite data indicate that the Yukon and Taku rivers’ watershed lakes are similar to each other as are the lakes within the Fraser River watershed. By contrast, the Mackenzie River watershed is more variable with Traviallant Lake being more similar to that of the Fraser River watershed and Frances Lake being more similar to the Yukon and Taku rivers.  The latter result is consistent with the geographic proximity between Frances Lake and the Yukon River.  Genetic divergence estimates among Fraser River watershed populations suggests a high level of genetic exchange during postglacial recolonization.  The signature of recent genetic exchange among Fraser River populations is consistent with the strong pattern of isolation-by-distance seen within this watershed but not between watersheds. Although the number of alleles across all populations and loci averaged 5.3 which is higher than that of 3.7 alleles per locus found in previous western lake trout study (Tamkee and Taylor 2003) the average in my study is perhaps skewed by the high degree of diversity in the two largest lakes in the study. In addition, it is difficult to compare across studies directly if the same loci are not   36 shared between studies.  The highest individual lake average for one population was found to be 11.25 (Tagish Lake) while the lowest was 3.13 (Airline Lake). The correlation between lake size and allelic richness is consistent with the study by Ihssen et al. (1988) who employed allozyme data and observed a significant trend towards greater diversity in larger lake trout lakes than in smaller ones. Another study (Tamkee and Taylor 2003) examined western lake trout in and around Banff, Alberta found an FST of 0.286 between two lakes from separate river systems compared to an FST of 0.212 over all my study lakes.  My study lakes comprised a mixture of within watershed and between watershed comparisons, which is not the same as pairwise analyses within a single watershed.    My study yielded FST’s that ranged from 0.014, which was found between two lakes (Atlin and Tagish lakes) found within the same watershed and connected by a nearby river,  to 0.616 between two lakes (Trapper and Bednesti lakes) separated by 780 km in different watersheds and colonized postglacially by separate refugia. Due to the unequal sampling across years, a temporal method of estimating effective population size was not possible.  Using the linkage disequilibrium (LD) method I was able to estimate and compare effective population sizes (Ne) in a relative manner and demonstrated that effective population size was correlated with lake size.   The benefits of the calculating Ne based on the LD method are that only year of sampling is necessary, however, the stability of the population is not known and, therefore, the Ne values generated should be considered with care. In some cases, lakes such as Pinchi Lake were estimated to have low Ne   37 yet allelic richness remained high which may be a product of the method employed or due to spawning specificity in the lake maintaining individual allelic uniqueness.  Caugley (1994) stated that genetic diversity can be maintained in natural populations with much smaller effective population sizes than expected. How small that Ne is in lake trout has, however, not been determined. The Ne estimates that I calculated do provide a basis for inferences about the potential genetic “health” of each lake and a baseline for future reference. Generally low Ne and low allelic richness were found in lakes with small sample sizes (e.g. Arctic Lake, Muncho Lake, Traviallant Lake) so some caution must be applied in assessing the genetic risks of potential low Ne in these lakes since the estimates may be a result of low sample size and not accurately reflect the true value. Atlin Lake vs Tagish Lake Atlin and Tagish lakes were the only two populations which displayed some deviations from HWE which suggested a degree of substructure might be present.  This suggestion was supported by the STRUCTURE analysis which indicated multiple subpopulations within Atlin and Tagish lakes.  STRUCTURE analysis indicated the presences of only two populations when analyzing data from all 19 study lakes yet indicate five and four populations for Atlin Lake and Tagish Lake, respectively. Western lake trout are comprised of two large populations based most likely on the refugia, as discussed earlier.  When these two populations are looked at individually, finer levels of subdivision may be identified; in this case, Atlin Lake and Tagish Lake are comprised of multiple subpopulations more closely related than populations from other lakes in the   38 region, and with continued gene flow between them.   Previous studies have reported genetic subdivision within single geographic localities among Arctic char, sockeye salmon, (summarized in Coyne and Orr 2004), as well as, lake trout (Ihssen et al. 1988, Page et al. 2003).  The Atlin and Tagish Lake populations were sampled over multiple years to account for interannual variation (Perkins et al. 1995) and, therefore, provide confidence that documented subdivision is stable over time.  Assuming that there are five and four subpopulations within Atlin and Tagish lakes, respectively, the level of subdivision found within Atlin Lake subpopulations (FST = 0.046) was greater than between Atlin and Tagish lakes (FST = 0.014).   Atlin Lake and Tagish Lake are connected by a river locally known as Atlin River which has been observed to contain lake trout (Mark Connor, Tlingit First Nation Fisheries Department, Atlin BC, personal communitcation).  Interlake dispersal via this river probably explains the close relatedness between some fish found in Atlin and Tagish lakes. Subpopulations within Atlin and Tagish lakes had lower levels of subdivision (FST = 0.011 - 0.084) than compared to lake trout in a study between populations within Lake Huron and Lake Michigan (FST = 0.012 - 0.142) (Page et al. 2003). These latter populations, however, have been repopulated from hatcheries geographically separated prior to their introduction into the lake perhaps allowing for an exaggeration of divergence to occur. An additional study by Piller et al. (2005) of other populations within the Laurentian Great Lakes reported lower levels of genetic differentiation (FST =   39 0.023 – 0.085), and Ihssen et al. (1988) found even lower levels within Lake Superior alone (FST = 0.018). These latter studies are more comparable to the range of values found in Atlin and Tagish lakes which suggests that sympatric populations have lower levels of subdivision.     Although five distinctive groups were found within Atlin Lake and four in Tagish Lake, many individuals were found to be highly admixed whose multilocus genotypes were comprised of 50% or less of any one population.  Although the STRUCTURE analyses, along with measures of subdivision, indicate the existence of subpopulations within both Atlin and Tagish lakes, it also considers the presence of admixture and further analysis suggests that this structure is only weakly developed. For instance, analysis of molecular variance showed that the majority of the variation found in these two lakes was within population variation, which agreed with the principal components analysis that although it indicated some clustering neither axis was significant.  When calculating the identity index of each subpopulation to compare relatedness of the subpopulations it was seen that all subpopulations had higher average relatedness within their own subpopulations, as expected. Some subpopulations, however, had similar values of within and between subpopulations relatedness as well (Table13).  Atlin Lake subpopulations’ relatedness agreed with the PCA clusters, i.e., the relatedness measures were all very close and neither axis was significant in the PCA.  For Tagish Lake the individual relatedness identities and PCA analysis were also consistent with one another and indicated weak yet significant differentiation within the lake. The relatedness values and the geographic positions within the lake, however, did not   40 always correspond.  In Atlin Lake subpopulations A and B were found to be the farthest apart geographically yet A and E were the most genetically distinct from one another.  Finally, the FST values were higher in Tagish Lake than in Atlin Lake perhaps suggesting that sampling from the spawning grounds yields a better estimate of the population structure. The patterns of relatedness between each of the subpopulations within the lakes may be explained by spawning activity. Both Ihssen et al. (1988) and Wilson and Hebert (1996) suggested a relationship between the genetic composition of a population and the geographic locations of source populations.  In their case the source populations were hatchery introductions where as in the wild the source populations now are the spawning shoals that colonized the lake from glacial refugia. Studies of both biochemical genetic variation and tagging data strongly suggest that lake trout have very precise homing, returning to their spawning site year-after-year (Ihssen et al. 1988).   These studies also indicated that the typical dispersal range of most lake trout is about 30 km while many tagged adults have been found at distances upwards of 50km from their spawning areas. Most adults, however, do not range more than 100km from their spawning shoals (Ihssen et al. 1988).   It has further been suspected that lake trout populations found on spawning shoals that are between 10 and 50km apart tend to utilize distinct cues for spawning bed use and that distance is not the sole isolating factor (Perkins et al. 1995).  In Atlin Lake, 86km of the shoreline were visually surveyed to identify seven probable spawning sites, as well as, seven suspected spawning sites all within the central core of the lake (Figure 10, Atlin Community   41 Working Group 2001).   The location of all spawning beds in close proximity to each other may be one of the reasons for such apparently high gene flow. Not all of the shoreline in Atlin Lake has been surveyed, however, and there may be other spawning shoals which have yet to be discovered.    Both in Tagish and Atlin lakes, spawning activity was found to be during the first few weeks of October. Therefore, any distinct temporal cues can be eliminated as a source for differentiation between the lakes. Whether or not the relatedness of fish from each subpopulation is driven by spawning site location within Atlin Lake cannot yet be determined from my data because samples were not collected from the spawning beds.  In Tagish Lake, as mentioned, portions of the samples were from the spawning beds.   Three spawning bed locations had sufficiently high sample numbers to determine if they were primarily from one population or not. The spawning bed (W) found the furthest north was found to be primarily genetic subpopulation TAG-B while the spawning bed (DB) in the southern portion of the lake was found to be composed almost equally of genetic subpopulations TAG-A and TAG-D.  The spawning bed found in the central portion of the lake showed roughly identical contributions from each subpopulation.  This distribution appears to be following similar patterns found in previous lakes with spawning beds further apart being composed of distinctive subpopulations.  The central (SW) spawning bed’s composition may be due to its location being in the dispersal range of lake trout from the other spawning beds as well as its own subpopulation.  Why, however, TAG-A and D were found on the same spawning bed yet were the least related, is not known.  Perhaps these two subpopulations   42 share spawning beds yet do not interbreed randomly.  Information on the morphology of each subpopulation may be useful to address this question. The low levels of subdivision in Atlin and Tagish lakes could stem from two phenomena:  (1) subpopulations are newly divergent and have yet to show isolation since postglacial recolonization, or (2) the contemporary subpopulations are remnants of previously more strongly structured populations that stem from postglacial recolonization from multiple refugia which is slowly becoming homogenized from high interlocality gene flow resulting in highly admixed subpopulations. I tend to favour the latter hypothesis given that there was evidence from the microsatellite data that these lakes were colonized from separate refugia.  However, the varied habitats and food sources available could support the maintenance of subpopulations, and this makes me hesitant to rule out the possibility of future subdivision.  In Atlin Lake gene flow between the subpopulations may be constraining natural selection and genetic drift from establishing and maintaining local genetic differences by spreading alleles and allele combinations resulting in a gradient rather than completely distinct subpopulations.  Until additional information on the spawning bed fidelity and genetic structuring related to the spawning beds is available it is premature to disregard either explanation. Conservation Implications Populations with a greater level of substructure are important to conservation because their higher degree of genetic divergence may signal the presence of key traits for adapting to changing conditions (Frankham et al. 2002).  My   43 evidence for the presence of multiple subpopulations within Atlin Lake along with its high allelic diversity and its close association with Tagish Lake make it a locale worth conserving.  This system is unique in that it is, to my knowledge, the only documented case of two highly diverse lakes connected by a well-used dispersal corridor in western Canada.  Lake Superior has small rivers extending from the main lake basin which, contain river spawning lake trout (Gunn et al. 2003).  No known spawning occurs in Atlin River; however, the frequent use of this river by lake trout I believe is one of the reasons high diversity is and will be maintained.  In order to maintain this diversity and perhaps the substructure spawning sites must be preserved.  The loss of spawning beds has been ranked high, along with over fishing and predations, as reasons for population decline. The Ne within the lake seems to indicate that population as a whole is quite large and diverse; however, some of the subpopulations are healthier than others with subpopulation A in Atlin Lake probably needing to be monitored closely in the future. Many of long held beliefs about lake trout ecology, which were based on the Laurentian Great Lakes populations, such as a narrow ecological niche and limited tolerance to temperature and oxygen levels have been dismissed as false (Wilson and Mandrak 2003).  If the Laurentian Great Lakes are quantitatively different, as some have suggested (Wilson and Mandrak 2003), then populations outside this area may hold the key to better understanding the biology of lake trout as a species.  My data on western lake trout will help to fill the void of information on the species across its range.   44 Conclusion The level of genetic diversity found in western lake trout agrees with previous analysis that suggests fish from multiple refugia repopulated these lakes following the last ice age.  Atlin Lake was found to have the highest degree of genetic diversity of all lakes sampled.  This was also true when compared to Laurentian Great Lakes lending weight to the theory that both Nahanni and Beringia provided postglacial colonists to the lake.    Atlin Lake was also found to contain some genetic structure; however, high gene flow between the subpopulations indicates these to be either developing or blending rather than completely isolated populations.   Additional information is required on the spawning shoals within Atlin Lake to determine if there are spawning beds in the northern and southern ends of the lake.  Secondly, additional sampling from the spawning beds is required to determine if there is a microspatial geographical isolation of some the subpopulations and to determine whether different spawning cues may be aiding in the subpopulation isolation within specific spawning shoals.     45 Chapter 3:  Phenotypic diversity as it compares to genetic diversity in Salvelinus namaycush in Atlin Lake, British Columbia Introduction Lake trout have long been noted for their variety in shape and colour (Gunn et al. 2003).  The Laurentian and Mackenzie Great Lakes are the only lakes that have been found thus far to possess sympatric morphotypes.  Although the Laurentian Great Lakes no longer possess the historical levels of diversity, (Brown et al. 1981; Burnham-Curtis and Smith 1994) they still contain three distinct morphotypes: the “lean”, “humper”, and “siscowet” forms. Each is recognized on the basis of facial characteristics, body fat content, spawning location and time (Brunham-Curtis and Smith 1994) and depth distribution (Moore and Bronte 2001). Each of these distinct morphotypes has been observed to occupy generally distinct habitats; however, it is not possible to differentiate individual fish solely on the basis of sampling locale (Dehring et al. 1981). Until recently very little was known about the phenotypic diversity of lake trout outside of the Laurentian Great Lakes and certainly no large lakes have been extensively studied.  The first documented case of sympatric morphotypes of lake trout outside of the Laurentian Great Lakes was in Great Bear Lake, NWT (Alfonso 2004).  It contains two morphotypes, one of which is similar to the lean form found in the Great Lakes.  Unlike the types found in the Great Lakes these two morphotypes are not spatially separate (Alfonso 2004).   Great Slave Lake   46 has also been reported to contain sympatric morphotypes (Zimmerman et al. 2006) as has Lake Mistassini, Quebec (Zimmerman et al. 2007).  Studies within Great Slave Lake have noted an association between body shape and water depth (Zimmerman et al. 2006).  The exact reason for morphological diversity within lake trout is not known, however, water depth is most certainly a factor (Zimmerman et al. 2006) and phenotypic plasticity, genetic drift, and fidelity to spawning areas have also been suggested as possible causes (Bronte and Moore 2007).  In Great Bear Lake an additional explanation has been proposed; that being resource polymorphism (Blackie et al. 2003). Consistently different spawning times have been observed between sympatric morphotypes of lake trout (Perkins et al. 1995).  Lower genetic diversity in lake trout genetic variability has been observed in small lakes compared to large lakes (Ihssen et al. 1988) as has a positive correlation between lake size and biomass (Hanson and Leggett 1982).  Their precise homing and site imprinting of lake trout, which is common to many salmonids (Ihssen et al. 1988; Hendry and Sterns 2004; Bronte and Moore 2007), coupled with the variety of possible spawning areas in large lakes probably promotes the formation of multiple distinctive populations. Despite the fact that lake trout appear to all spawn at a similar time of year in Atlin Lake (Atlin Community Fisheries Working Group 2001) there are other factors present such as variation in food source, water depth, water turbidity and lake size, which may allow for local differentiation to occur.  Differences in colouration and body shape have been observed within Atlin Lake.  Whether the observed morphological variation can be attributed to   47 local adaptation or to purely environmental variation is of great interest to biologists and conservationists alike.  The relationship between phenotypic variation, genetics and the fitness consequences of such variation needs to be determined to understand the diversification of species within ecosystems (e.g., Keeley et al. 2007). Due to the high levels of morphological variation and the degree of overlap between characteristics it becomes difficult to identify all lake trout as a specific morphotype (Dehring et al. 1981; Brunham-Curtis 1993; Brunham-Curtis and Smith 1994; Moore and Bronte 2001; Bronte and Moore 2007).  Initial studies of phenotype suggested a genetic component to morphological differences and the frequency of mixed phenotypes suggests gene flow among the phenotypes (Brunham- Curtis and Smith 1994; Page et al. 2003). Areas with different frequencies of each phenotype have been distinguished based on allelic frequency differences and the presence or absence of specific alleles (Ihssen et al. 1988).  Allozyme data also indicated some geographic population structuring within the morphotypes (Dehring et al. 1981; Ihssen et al. 1988) and microsatellite frequency data revealed genetic differences between the three morphotypes in Lake Superior (Page et al. 2003).  It has been long suspected that there is an association between genetic and phenotypic variation, but as of yet no study outside the Laurentian Great Lakes has been performed with lake trout to verify this hypothesis more broadly. If Atlin Lake is found to possess more than one discrete class of morphotype than the next step is to determine if each is genetically distinguishable.  Such genetically-based morphological variation is   48 important to document as it suggests that local adaptation is an important process in promoting and maintaining a diversity of populations within a single species (Keeley et al. 2007). The degree of concordance between molecular genetic and morphological structuring has been variable.   For instance, both historical (Kruger and Ihssen 1995) and contemporary (Page et al. 2003) data indicate that multiple lake trout ‘stocks’ existed within the Laurentian Great Lakes.  While morphological data from Lake Superior has been found to cluster into three main geographical areas with finer scale structure being suggested; it is thought that there might be upwards of ten different genetic populations but three common morphotypes (Bronte and Moore 2007) suggesting that the degree of genetic substructure in lake trout is greater than phenotypic substructure.    Whether such a case prevails in Atlin Lake is of note for two reasons.  First, if multiple genetic and morphological populations exist in Atlin Lake, management plans will have to be designed to consider both genetic and morphological levels of subpopulation. Secondly, if the degrees of genetic and morphological population structure observed do not agree, understanding the causes of such variation could provide valuable insights into the biology of lake trout particularly in terms of habitat and feeding biology. Objective Past studies have identified lake trout populations in a variety of ways; through fat content, swim bladder, development rate of eggs, survival, depth distributions, reproduction, morphology, and through the use of genetic tools such as   49 allozymes, mitochondrial DNA, karyotypes, and microsatellites (Krueger et al. 1989). Rarely have the results of the multiple methods been compared to determine if the phenotypic and the genetic data of the same individuals correlate with one another.  In this chapter I compare the results of the genetic structure analysis which I conducted in Chapter 2 with the morphological data collected in this chapter to address the following questions: (1) Are there multiple distinct morphotypes within Atlin Lake?  (2) Does the genetic population structure agree with the phenotypic structure observed in the lake?  Such information would not only help answer many of the general questions lake trout biologists have posed, but also aid in  developing an effective management strategy for Atlin Lake. Visual identification of each sub-population would ease monitoring capabilities by reducing the cost and time necessary to determine which subpopulations are where and in what numbers using more intensive genetic analyses.  Materials & Methods Sample collection and DNA analysis see Chapter 2 Morphological Measurements Photographs of individual fish were taken using a digital camera with the measuring board as a consistent background for light standardization. Body colouration, or more accurately brightness, of each lake trout image was measured with the aid of Image J version 1.32j software (http://rsb.info.nih.gov/ij) following the methodology of Zimmerman et al. (2006).  All images were   50 converted to black and white for ease of interpretation.  Six measurements were taken of each image (Figure 11) on a 0 (black) to 256 (white) scale and averaged for a total overall “brightness factor”. Body shape was measured with the aid of TPSDIG software (Rolph 1997; http://life.bio.sunysb.edu/morph).   The digital images were imported into the program and x and y coordinates were established using 16 landmarks to measure the shape of each fish (Figure 12 Zimmerman et al. 2006; Moore and Bronte 2001).  These data were then imported into a companion program TPSRELW which centers, scales and aligns the coordinates using the Procrustes method and calculates a consensus configuration to describe the overall body shape from a sample of fish.  The TPSDIG digital software was further employed to measure head depth, mid-body depth, caudal peduncle depth and confirm body length which had been measured in the field (Figure 13). These latter measurements in combination with the landmark measurements resulted in a total of 36 body shape measurements. To account for variation in body size- related differences, individuals were size corrected to a common fork length and the residuals from the slope employed in statistical tests rather than direct measures. Data Analysis A principal components analysis (PCA) was performed using the correlation matrix among the 36 morphological measures, including 32 aligned body shape measures, with the aid of the statistical program PAST (Hammer et al.  2001).   51 To determine the number of morphological clusters which could be distinguished a model based clustering method was employed without a priori designation of populations using the program MCLUST (Fraley and Raftery 2003; implemented in R).  The model describing K populations with the highest Bayesian information was selected for subsequent analysis. A contingency test was also performed using the program PAST to determine (1) if the morphologies were associated with geography and (2) if there is a relationship between the genetic subpopulations determined in Chapter 2 and the morphological clusters. A discriminant function analysis (DFA) was performed in PAST to measure the reliability of assignment based on the 36 morphological characters employed in the cluster analysis. Due to the observations of geographic colouration this morphological feature was tested as a single variable as well as in combination with all the other morphological characters in the PCA.  ANOVA’s were performed to assess the level of colour differentiation among (1) the genetic subpopulations, (2) morphological clusters, and (3) geographical units.  Results Data were collected and analyzed for 104 lake trout. Four individuals were photographed and analyzed twice to determine repeatability of analysis.   Of these four samples three were assigned to the same morphological group each time they were measured while the fourth assigned to one group 50% of the time   52 and the other the remainder suggesting high repeatability of any morphological differences observed. Two components in the PCA accounted for 56.2% of total variation.  Principal components 1 and 2 (PC1 and PC2) made up 40.7% and 15.5% of the total variation, respectively. PC1 represented a contrast between mid positive loadings for coordinates y6, y7, y13 and y14 which represent body depth and fin position and low negative loadings for coordinates x5 which represent the curvature of the head along with y9, y10, and y11 which correspond to body length.  In PC2, head shape had the highest loadings (x1, x3, x4).  PC2 had low negative values for head depth, mid-body length, caudal peduncle width and condition factor compared to a positive value for each of these in PC1; however, neither PC had strong values for these characters (Table 19). Two distinct clusters were resolved following the criterion of multiple models (Figure 15) with one cluster having a greater degree of variance than the other (Figure 16).  Morphological populations were found to be weakly geographical distributed as the contingency test yielded a statistically significant difference (P = 0.046) in the proportions of the two morphotypes among the different geographical units (Table 20; Figure 17). An additional contingency test failed to demonstrate statistical differences (P = 0.35) in morphotype composition among the five genetic subpopulations identified in Chapter 2 (Table 21). Morphological population 1 was distributed relatively uniformly across all genetic subpopulations, however,  morphological population 2 was comprised of a larger fraction of  genetic subpopulations B and E; however, it also possessed a   53 proportion of subpopulations C and D.   The DFA showed a 75.9% correct morphological assignment with a third of the miss-assignments occurring by fish from morphology group 2 assigned to group 1  (Figure 18). Colouration was found not to be statistically different between the genetic subpopulations (P = 0.13) or the morphological clusters (P = 0.15).  The observed geographical distribution of colour was, however, found to be significant (P = 0.016) with the lighter variety being more common in the south (Figure 19). An FST of 0.004 (95% confidence interval of -0.001 to 0.009) was calculated between the two morphological clusters which was an order of magnitude lower than the value of the global FST of 0.046 (0.025-0.086) among the five genetic subpopulations determined in Chapter 2.  Discussion This study illustrates the presence of two morphological kinds of lake trout within Altin Lake; two morphotypes that appear to differ primarily in head shape and fin position.  The three morphotypes found within the Laurentian Great Lakes are known to have significant differences in head shape (Moore and Bronte 2001). The Great Bear Lake “redfin” morphotype was also noted to have smaller head and dorsal fin position than the “normal” lake trout (Alfonso 2004).  Multiple phenotypes have been reported from Atlin Lake by locals for generations, but whether the different morphotypes were discrete units or a continuous range of   54 variation was not known (M. Connor, Tlingit First Nation Fisheries Department, Atlin BC, personal communication). The obvious differences in colouration that have been observed within Atlin Lake could be due to unique genetic characteristics; however, there was an association found between geographic location and colour.  It has been noted in other salmonids that flesh colour is related to diet (Brown et al. 1981) as is external colour (Scott and Crossman 1973).  Zimmerman et al. (2006, 2007) found colouration differences to be associated with different body forms that were found at different water depths.   The extreme colouration differences (Figure 19) that I observed throughout the lake are most likely due to different environments, principally in terms of turbidity and possibly water colour.   It is interesting to note that the conditions found at the southern portion of Atlin Lake are presumably most similar to those of the proglacial lakes following the Wisconsian glaciation (Power 2002).  The high level of turbidity caused by the influx of sediment from glacial meltwater might be a cause for the distinctive silver colouration of the lake trout found in this area.  In fact, locally these fish are known as the “glacial” variety. All fish captured in the lake survey were found in depths less than 50m. The increased turbidy caused by the glacial meltwater in the south end of the lake may be simulating the conditions in the deeper waters.  The colouration differences found in Lake Mistassini (Zimmerman et al. 2007) showed individuals to be lighter in the deeper waters than those in shallow water. The matter is complicated, however, by the fact that lake trout in Great Slave Lake (Zimmerman et al. 2006) exhibited the reverse pattern (i.e. light in shallow water   55 and dark in deep).  Perhaps if fish in Atlin Lake were caught in deeper water an association between depth and colour might also be found.   Interestingly, geographic segregation of colour morphs of lake trout has also been reported in Great Bear Lake (i.e. silver in the south, dark in the north), but the basis of this pattern in unknown. Despite the identification of two morphotypes from my analysis, there was considerable overlap between the types, which was confirmed by approximately 30% level of miss-assignment.  Due to the overlap and subtle differences between the morphotypes it was difficult to assign an individual fish to one type or the other in the field by visual identification.   This presence of two morphotypes and subtle differentiation could, in part, be driven by sexual dimorphism. Identification of sex in lake trout can not be determined through the non lethal sampling I preformed, therefore, although unlikely, I am unable to dismiss the possibility that the two morphotypes represent males and females. There are, however, no reported cases of sexual dimorphism in lake trout; excluding during the spawning season.  In addition, the colouration differences are unlikely to be explained by sexual dimorphism given the clear association of colour types with lake colour and associated crypsis. Finally, as the morphotypes showed significant geographic associations, if the morphotypes were driven by sexual dimorphism, one would need to envoke sexual differences in habitat use to explain the association between morphotype and geography.    56 The morphological populations resolved did not correspond closely to the subpopulations defined through microsatellite analysis (Chapter 2).  Although genetic subpopulation A was found to occur only in morphological population 1 there were only 4 samples which were able to be used in the McCLUST analysis. Morphological population 1 also contained fish from all of the other four genetic subpopulations, which contained the greatest number of samples yet had the smallest variance.   By contrast, genetic subpopulations B and E made the largest contribution to morphological population 2.    In addition, my analyses indicated that there was no significant genetic differentiation between the morphotypes, as least as measured using microsatellites.  This latter result could result from:  (1) Morphological variation of the characters studied is primarily due to environmental differences among localities and members of different genetic subpopulations use the same environments during non-spawning periods, or (2) The genetic subpopulations represent selectively neutral differences that move among adaptively-differentiated morphological populations via some gene flow. I have limited evidence to support the first of these conclusions; however, others have speculated about the genetic versus environmental control of phenotype in lake trout.  Evidence suggests that lake trout historically evolved and diverged from other Salvelinus species in response to environmental changes caused by the Pleistocene glaciations (Wilson and Mandrak 2003). The three main morphotypes of lake trout found in the Laurentian Great Lakes differ significantly in microsatellite allele frequency (Page et al. 2004) and studies since have made the assumption that a genetic component underlies the morphological differences   57 present (Bronte and Moore 2007).   As yet no genetic information has been provided for the other morphologically diverse lake trout in the Mackenzie Great Lakes. Other salmonids have demonstrated morphological variation due to environmental effects such as brook trout, Salvelinus fontinalis (Imre et al. 2002) and Atlantic salmon, Salmo salar (Pakkasmaa et al. 2001).  Both demonstrated morphological change in body shape and fin size due to water velocity.  In Arctic char, Salvelinus alpinus, Adams et al. (2003) found the rearing environment was the primary cause of morphological variation; however, in this case, there was also a small genetic component.   A common garden experiment could be performed with Atlin Lake lake trout to test the genetic basis of the phenotypic variation observed.  By selecting lake trout from each of the geographic units, breeding them, exposing their offspring to different and identical environmental conditions and measuring the morphological features environmental control vs. genetic control could be assessed.   Lake trout grown in hatcheries have been known to maintain their Laurentian parental morphology suggesting a genetic component involved (Moore and Bronte 2001), but whether that would be the case of Atlin Lake lake trout is uncertain. Differences in spawning fidelity have been noted between lakes and between different stocks of the same morphology within a lake.  Wild lake trout have a much higher fidelity than hatchery-reared fish suggesting some form of local adaptation (Wilson and Mandrak 2003; Bronte et al. 2007).  The differences in fidelity to spawning localities could be dependent on morphotype, and, therefore, some morphotypes may promote gene flow among genetic populations more   58 than others.    Hendry and Sterns (2004) discussed how divergent adaptation and homing in salmonids reduces the fitness of strays.  They further mention that most fish stray to nearby spawning sites, therefore, genetic differentiation is most often positively correlated with geographic distance.  The fact that spawning sites are close to one another in Atlin Lake may be the reason why an association between genetic and morphological variation has not been found.  Despite the abundant anecdotal data on lake trout populations there is relatively little evidence published concerning adaptive variation.  Hendry and Taylor (2004) found that gene flow did have substantial effects on adaptive divergence in sticklebacks; however, the degree to which it has an effect depends on the trait. Further genetic and morphological information from the individuals at each of the spawning grounds is necessary to address some of the remaining questions. Lake trout morphotypes in the Laurentian Great Lakes vary in growth, fat content, depth preference, age of maturity, mortality, diet, spawning time, as well as location and body shape (Moore and Bronte 2001). Assignment tests based on snout shape, eye size and body shape are insufficient to distinguish between the morphotypes due to the high degree of overlap. There was no apparent association between the established morphotypes from the Laurentian Great Lakes and those found in Atlin Lake, but the data are really too few for a definitive analysis. My casual visual identifications based on the morphological descriptions suggest, however, that, indeed, Atlin Lake contains both a lean and a siscowet variety, unfortunately, none of the siscowet type were included in the morphological analysis due to poor quality photographs.   A humper type was   59 also observed, however, I was unable to determine if it was an additional morphotype or would have clustered with the other siscowet type observed due to the lack of siscowet samples.   Further morphological and genetic study of lake trout should be based on determining if the classic morphotypes reported from the  Laurentian and Mackenzie Great Lakes, such as the siscowet, are the same morphologically as those in Atlin Lake and if such similarity results from parallel evolution.    Parallel evolution has been noted in northern freshwater fishes (Behnke 1972; Taylor 1999) such as sticklebacks (Schluter et al. 2004) and whitefish (Østbye et al. 2006).   Lake trout are freshwater fish and as such have been separated by watershed boundaries for thousands of years isolating populations each of which experienced different environmental conditions (Keeley et al. 2006). By comparing the conditions of the lakes and the morphotypes present, a greater appreciation of the sources of variation and adaptability of the species to different swimming requirements, food resources, as well as, lake water colour for prey protection could be obtained. Conclusion Two distinct morphological populations were found within Atlin Lake, but these morphotypes were not genetically distinguishable.   The degree of environmental impact on the morphology is uncertain; however, body colouration does appear to be geographically associated and is suspected to result from lake-wide differences in glacial-influenced turbidity levels.    Due to some remaining uncertainty and the fact that the lake was sampled unevenly, additional study of morphological variation in Atlin Lake lake trout is warranted.   If the morphological   60 and genetic variation is found, on further study, to be correlated this would affect the management plan by having a visual identifier for part of the genetic subdivision present.  Both commercial and recreational fishing would be able to more easily monitor the increase or decrease of either morphology to determine which genetic subpopulations were suffering before significant depletion had occurred.   If the morphotypes remain distinctive from the genetic structure than the habitats and environmental features allowing for such division would have an increased value in terms of conservation and provide further insight into the biology of lake trout.    61 Chapter 4:  Contribution of Salvelinus namaycush populations in Atlin Lake to local fisheries Introduction A common problem in fisheries management is that sometimes it involves only the management of fish when it should include assessment of fishers and their effort (Hilbron 2007). More enlightened resource management involves plans established that are appropriate for the species and area, as well as being feasible for the fishers.  In striving for such a plan for lake trout in Atlin Lake, BC, basic knowledge about species biology (Chapter 2 and 3) as well as how the current fishing practices are effecting the resident lake trout populations is essential. Declines in abundance of populations can lead to overall species declines in morphology and genetic diversity (see Chapter 2).  Freshwater fish species, especially, are at high risk due to invasions by nonnative species and human activities (Guinand et al. 2003).   It is the risk caused by human activities that we seek to minimize by developing appropriate and workable management plans. Baseline levels of genetic diversity (Chapter 2) and morphological diversity (Chapter 3) need to be established before we can determine how present or future fisheries may affect Atlin Lake lake trout diversity and persistence. Lake trout continue to be an important source of food, revenue, and recreation for communities particularly in Atlin Lake. Consequently, there is great interest in conservation and, where necessary, recovery of the species (Atlin Area   62 Community Fisheries Working Group 2001).  There have been numerous studies whose main goal has been to determine the best method of revitalizing diminishing populations throughout North America (Kruger et al. 1989; Guinand et al. 2003). It was at one time thought that no other lake outside the Laurentian Great Lakes possessed multiple lake trout morphotypes.   That perception has since been proved inaccurate (Alfonso 2002, Zimmerman 2006, Chapter 3). Each lake possessing multiple morphotypes and lakes with high diversity are of great conservation value as they help protect the long term viability by conserving the evolutionary and adaptive potential for the entire species (Piller et al. 2005). Due to the substructure found within Atlin Lake that I have uncovered, it is important to assess the degree to which each subpopulation may contribute to the various fisheries.    Fisheries that might target less productive subpopulations could, for instance, reduce the level of diversity in the lake may put them at risk and, therefore, compromise the persistence of diversity within the lake as a whole. However, such loss of diversity does not by itself imply loss in overall fishery production over the short term, but rather loss in potential yield should circumstances change so as to favor subpopulations that have been lost. It is potentially valuable when managing an area with multiple populations that each of the populations be identified and their contributions to the fisheries determined (Larkin 1981 cited by Ryman and Utter 1987) to be certain no one population is contributing more than the others, or if it is, that such populations can sustain high levels of exploitation.   If one population is contributing more   63 there is a possibility of reducing morphological and genetic diversity or inducing demographic collapse through overexploitation particularly of small populations. Without accurate information about the populations that occupy a system it is difficult to determine if the less productive populations are being protected or the more productive populations are being under harvested (Pella and Milner 1987). As previously mentioned, subdivision within Atlin Lake is based on the significant level of genetic diversity between groups; however, it is unknown what proportion each of these subpopulations contributes to the fisheries.  Previous studies on bull trout in coastal BC (Taylor and Costello 2006) and lake trout in the Laurentian Great Lakes (DeKoning et al. 2006) have employed a method called genetic mixture analysis to address a similar question.  The objective is to compare each fishery sample to a baseline consisting of all subpopulations and determine the proportional contribution of each subpopulation to the mixture.  By identifying the various contributions of each subpopulation, these contributions can be monitored, creating the possibility of allowing the harvest to be managed at a level appropriate for the productivity of the weaker subpopulation as a risk management approach. This approach has a long history especially in commercial salmon fisheries management (Pella and Miller 1987; Larsen et al. 2005; Taylor and Costello 2006) along with being used for estimating the contribution of hatchery strains of lake trout to Laurentian Great Lakes’ fisheries (Angers and Bernatchez 1996; Page et al. 2003; DeKoning et al. 2006).     64 Objective In this chapter I examine the proportional contributions of the genetic subpopulations found within Atlin Lake to the fisheries. Microsatellite markers along with mixed stock analysis were employed to differentiate lake trout and to generate estimates of proportional contributions to commercial and recreational fishery samples from Atlin Lake. Due to the location of the commercial fisheries and of the most common angling locations within the lake I suspected that subpopulations were not equally contributing to the fisheries and I sought to test this idea.  Materials & Methods Sample collection and DNA analysis Fin clipped samples were collected from Great Northern Fish Company (GNFC) and from local anglers and stored in 95% ethanol.  Each of these samples were genotyped following the same methodology as described in Chapter 2. Population genetic and mixture analyses Reference population data was determined using STRUCTURE version 2.1 (Pritchard et al. 2000) and GENECLASS version 2.0 (Piry et al. 2004).  As mentioned in Chapter 2, STRUCTURE identified the K populations for both Atlin and Tagish lakes’ survey samples.  The program further assigned each individual to its most likely population.  Survey samples were categorized and then self- assigned using the program GENECLASS.  GENECLASS employs the Bayesian   65 assignment methodology of Rannala and Mountain (1997).  This method was selected based on the recommendations of Cornuet et al. (1999).  It calculates the allele frequencies of all reference populations and determines the likelihood of each individual multilocus genotype occurring in each reference population taking into account the sampling error associated with estimating allele frequencies.   Monto-Carlo resampling followed the algorithm of Cornuet et al. (1999) with 10000 simulated individuals and an alpha value of 0.05.  These reference populations were used in conjunction with genetic admixture analysis to determine if the commercial fisheries in Atlin Lake were sampling proportionately from each population. To determine the contribution of each population to the fisheries I used the Bayesian approach to mixed stock analysis employed by the program GMA, Genetic Mixture Analysis (Kalinowski 2003). It utilizes the Rannala and Mountain (1997) method for estimating the probability of observing a genotype in a population each of which have been characterized across a number of genetic loci and constitute the “baseline” samples.  The program then estimates the proportion of each baseline population’s contribution to the mixed (fishery) sample (Kalinowski 2003). Sample sizes were equalized by simulating mixtures of 50 individuals by random sampling with replacement and estimating mixture proportions for 5000 replicate analyses.   In this way I was able to determine the robustness of this method by viewing the variability of mixture proportions across the replicates.   66 This program was initially employed to compare the commercial and angling samples to all 19 lakes sampled to determine the accuracy of this methodology and as a form of positive control. In this analysis, I expected that the mixture analysis would identify no non-Atlin/Tagish lakes as contributing to the Atlin Lake fisheries.  Next commercial and angling samples were run both together and individually comparing them to the sub-populations from both Atlin and Tagish lakes.  Initial reference populations were defined by assigning each sample collected to its most likely population including individuals that had a high proportion of alleles from other subpopulations.  Reference populations for both Atlin and Tagish lakes were further refined by removing samples which had admixed genomes (i.e., > 0.5 admixed).  Results Initial comparisons of lake trout commercial and angling samples (N=134) from Atlin Lake to all 19 lakes in this study indicated that 79.6% of these samples were from Atlin Lake while 20.4% were originated from Tagish Lake.   No other lakes were identified as contributing to the Atlin Lake fisheries. Lake trout from commercial fisheries and those caught by local anglers were compared to the subpopulations determined in Chapter 2 of both Atlin and Tagish lakes with the results being similar to whether or not individuals with > 0.5 admixed genomes were included (Table 22) or excluded (Table 23).  Commercial fisheries seem to draw most heavily from subpopulations Atlin B, C, E and Tagish A,C while the anglers drew more heavily from Atlin A, C,  as well as   67 Tagish D.   Atlin B also contributed to the angling catch, but to a lesser degree than to the commercial fisheries.  Two of the Tagish subpopulations (B and C) were not sampled at all by the anglers. Simulations were run to determine the accuracy of the mixture results.  The simulated mixtures showed similar results to that of the original mixture.  The most variation was found to occur in the subpopulations with the lowest contributions; for commercial samples subpopulations Atlin subpopulations D and E and angling Tagish subpopulations B and C (Tables 22 and 23).  No one subpopulation increased or decreased its contribution by any greater than two percent and in most cases far less.  Discussion Lake trout are important to humans for food and recreational angling, as well as being a key indicator species for the health of a lake.   It is important for conservation to: (1) identify population structure to quantify conservation units (Chapter 2) and (2) realize the impacts of current fishing practices on populations.  This second aspect of conservation is the focus of this chapter. Previous chapters of my thesis have shown that not only are lake trout diverse across the range (i.e., allelic richness and number of alleles from the Fraser to Mackenzie rivers), but also that Atlin Lake itself contains a diverse population of lake trout both genetically and phenotypically.  To determine the contribution of   68 each subpopulation to the fisheries, samples were collected from both recreational and commercial fishers to determine their genetic composition. Proportional contributions It is not surprising that lake trout from Atlin and Tagish lakes were the only individuals contributing to the fisheries and that fish from other lakes had zero estimated contribution.    Atlin/Tagish lakes have been isolated from all the other lakes for the past 10,000 years (Wilson and Hebert 1998) and thus have distinctive genotypes as compared to other western lake trout lakes.  Further, the fact that a portion (20.4%) of the samples collected from Atlin Lake fisheries was estimated to be from Tagish Lake is mostly likely due to the observed migration of lake trout through Atlin River which connects Atlin and Tagish lakes (Figure 13). Lake trout caught through angling have a slightly lower proportional contribution from Tagish Lake than do those from the commercial fisheries.  This seems reasonable because the Atlin River is closer to the commercial fisheries locations than areas that receive the greatest angling effort.   The simulated mixtures reflect the results of the mixture analysis demonstrating the robustness of my results.  The majority of the variability was found to be associated with the small contributors, which is not unexpected since it is difficult to estimate rare events and small changes can dramatically change the percentage contribution. Traditional mixed stock analysis assumes that fish from a given fishery are composed of a mixture of fish from several populations each of which are in Hardy-Weinburg equilibrium (Choisy et al. 2004, DeKoning et al. 2006).      The accuracy of individual assignment depends on the number of loci assayed, the   69 level of variability of genetic loci and the extent of divergence among source populations (Cornuet et al. 1999).  Due to the lower level of subdivision found within Atlin and Tagish lakes the individual assignment accuracy of reference populations employed in the mixture analysis was low providing a source of uncertainty in this analysis. To minimize this error the individuals without a minimum of 0.5 probability of assignment to one population were removed. Despite the lower sample sizes the mixture and simulations changed only slightly.  This more conservative reference population dataset and analysis showed that subpopulation C had the greatest contribution to the fisheries as a whole. The greater contribution of subpopulation C may be due to the larger estimated effective population size, which may in turn reflect a higher census size as well. The relative numbers of fish sampled from each population did not correspond directly to the estimated effective population size nor did those subpopulations more heavily sampled necessarily contribution greater to the fisheries.  For instance, overall Atlin Lake subpopulations B and C were most frequently sampled by commercial and angler fishermen as estimated by the genetic mixture analyses.  These two populations also ranked first and third for sample size in our collection survey as well as having the highest estimated effective population sizes (Chapter 2).  Atlin Lake subpopulations A and C, however, possessed the most number of unique alleles perhaps increasing their value as conservation units.   Subpopulation B in Tagish Lake possessed the largest effective population size and yet contributed the least to Atlin Lake’s fisheries. Its   70 contribution to fisheries in Tagish Lake is unknown due to the lack of samples from fisheries in Tagish Lake.  My results indicate that some subpopulations appear to contribute disproportionately to fisheries in Atlin Lake. The exact reasons why this might be are unclear, but may have to do with the proximity of fishery areas to main feeding or reproductive habitats of those populations.   The known spawning locations are found in the central portion of the lake (Figure 10) as are the more common angling locations (Figure 20); therefore, if the spawning locations correspond to specific spawning beds those subpopulations closer to popular angling locations could show a higher contribution. More information, such as which subpopulations visit each of the spawning sites, is required before this can hypothesis can be tested. Conservation implications Lake trout are sensitive to anthropogenic influence as seen by the dramatic reduction in diversity and population size found in the Laurentian Great Lakes due to over-fishing and the introduction of the sea lamprey (Gunn et al. 2003). The genetic and morphological subdivision within Atlin Lake makes the individual subpopulations of lake trout more susceptible to disturbance because the loss of a few adults may remove a large proportion from small subpopulation.  Large and subdivided populations, however, may also fare better under environmental disturbances and population declines than isolated populations because a source of migrants to depressed populations is possible which is not available for single isolated populations (Ihssen et al. 1988).   Due to the apparently high gene flow within Atlin Lake and between Atlin and Tagish lakes the risk of removing unique   71 gene complexes completely is lessened; however, the potential for negative genetic consequences of overexploitation is still present.  If gene flow among the subpopulations decreases, perhaps owing to environmental degradation in intervening habitats, more isolation could develop among subpopulations. My results indicate that there is genetic diversity within Atlin and Tagish lakes and that this diversity is unequally exploited in commercial and recreational fisheries. This suggests that in order for the lakes’ lake trout to be managed effectively, these differential contributions need to be acknowledged in management plans. For instance, if subpopulation B is, indeed, the most significant contributor to fisheries as my results suggest, a better understanding of the geographical and temporal distribution of these fish is a conservation priority. Conversely, populations that seem to contribute little may signal low productivity or low susceptibility to fishing. Knowledge about either of these alternatives would improve management of the resource. Finally, there are also fisheries in Tagish Lake, yet we have no information on the relative contributions of its subpopulations (and potentially those in Atlin Lake) to the Tagish Lake fishery. Information on the Tagish Lake fisheries is a priority to better managed the interconnected Atlin-Tagish watershed as a whole.     72 Chapter 5: General Discussion  Population structure of Atlin Lake Lake Trout My thesis found high levels of genetic diversity in western lake trout populations and that patterns of molecular genetic diversity in the Atlin Lake ecosystem suggested that the system was colonized by fish from multiple Pleistocene glacial refugia. Furthermore, substructuring of populations was found both in Atlin Lake and Tagish Lake.      Natural selection is unlikely to have influenced the genetic results observed because although some microsatellite may be linked to adaptive loci (Petit et al. 1998), in general microsatellites probably reflect neutral divergence.  Phenotypic vs Genetic Diversity Morphological variation was also present within Atlin Lake and fish were found to cluster into two distinct groups.   My thesis found the morphological and genetic subpopulations were not correlated with each other; however, morphology did appear to have a weak association with geographic location suggesting environment plays a key role in morphological differentiation. An important aspect of morphological variation is whether that variation is associated with performance differences of fish in nature.   The recognizable morphological difference between habitats involved the colouration of the fish with the silver variety found in more turbid waters while clear water tended to possess darker variety.  This pattern maybe a result of selection for predator avoidance; the   73 darker colour being less noticeable to birds of prey in the clear water while turbid water conditions mask the brighter silver variety from view (Endler 1987). The difference between wild lake trout and hatchery strains seems to indicate that there is some locally adaptive advantage in the Laurentian Great Lakes (Grewe et al. 1994; Perkins et al. 1995).  Whether this is the case in Atlin Lake is currently unknown but the observed colour patterns suggest this possibility. Causes of variation in other lakes have been speculated to be genetic drift, spawning fidelity and local adaptation (Bronte and Moore 2007).  A comparison between the causes of population divergence, both genetically and morphologically, in the all diverse lake trout lakes would be of great interest and give greater understanding to lake trout biology and evolution. Atlin Lake is the smallest lake known thus far to have multiple morphotypes and genetic populations; this may be due to the connectivity with its neighbor, Tagish Lake and the increasing habitat complexity that this represents. Whitlock (2003) stated that the degree of substructure within a population affects the time to fixation or loss of alleles, as well as, which alleles are more likely to fix. Greater amounts of substructure increase the overall Ne of the lake population, thereby, reducing the effects of drift in fixing deleterious alleles while providing selection with the ability to increase the frequency of advantageous alleles.  Following this theory, Atlin Lake and Tagish Lake’s lake trout substructure increases the chance of fixing beneficial alleles while decreasing the likelihood of deleterious alleles becoming fixed across the ecosystem as a whole. Atlin Lake further benefits from the migration of lake trout from Tagish Lake into   74 Atlin Lake, as was observed in the genetic mixture analysis because such interconnection may introduce variation from one area into another.   Tamkee (2006) suggested that the degree of connectiveness among rainbow trout populations influenced the effective population size at specific localities because of gene flow from individuals from nearby populations that increased genetic diversity of the system.   Consequently, in order to maintain fisheries habitat, a manger must consider Tagish Lake as well as Atlin Lake when making decisions. The lack of correlation between genetic and morphological structure demonstrates the necessity of not basing decisions solely on genetics or morphology, but to consider multiple aspects of a species’ variation. I suggest that there is no real fear of outbreeding depression due to high gene flow between the subpopulations.  If the phenotypic diversity present is due to local adaptation perhaps this lake system demonstrates a system that is in a kind of selection-gene flow balance with increased fitness due to local adaptation and multiple compatible subpopulations as sources of migrants to better withstand subpopulation disturbance. The presence of local adaptation within a species highlights the necessity in conserving intraspecific diversity which is necessary for populations to persist in ever changing habitats.  Fisheries Management It has been suggested that lake trout are best managed based on the goal of maintaining resource sustainability rather than strictly on harvest levels   (Wilson and Mandrak 2003) which I believe to be true.  Besides the fishing pressure   75 imposed by humans the second greatest concern due to anthropogenic forces is the disturbance of lake trout’s spawning grounds.  Presently there is an incomplete picture of the spawning grounds within Atlin Lake and no information on how the substructure identified corresponds to each location.  In order to effectively manage this lake system the remaining portion of the lake should be examined to determine if spawning shoals are present other than in the central section of the lake.  Knowing if there is a correlation between spawning shoal and subpopulation presence in Atlin Lake, as is the case in Tagish Lake (Chapter 2), would aid in determining which spawning areas are at higher risk. Despite the unknown reasons for the variation present the preservation of spawning beds and food resources within the lake are necessary to maintain the health of this ecosystem. At present a large proportion of Atlin Lake is protected under the provincial parks act and the majority of the remaining sections of the lake are uninhabited, however, the concentrated commercial fishing in the northern end must be monitored in order to ensure that the species does not become overtaxed by fishing.  No phenotypic information was gathered on the commercially caught fish allowing for the possibility that although the fisheries appear to be sampling proportionately from each genetic population there could be a disproportionate sampling of phenotypes.  If such morphological variation is or has led to local adaptation, preservation of such variation is required for the health and adaptability of the species to be maintained.   If stronger genetic substructure were to develop (Chapter 2) then an association might occur between the   76 morphotypes and genotypes as has been found in the Laurentian Great Lakes (Page et al. 2003).  Further study is needed to investigate the degree to which the different morphotypes contribute to the fisheries.  Conservation My thesis illustrates the great variability present within lake trout in the Atlin Lake system, which probably enhances the system’s evolutionary potential.  The phenotypic differences observed within the lake are most likely a result of postglacial local adaptation rather than marking distinct refugial origins.  Variation in local habitats may allow for each of the subpopulations to express and maintain their phenotypic differences and should be explored.  Whether these phenotypic differences have a resulting change in fitness and/or the degree to which they are genetically based is something that has yet to be determined. Although the phenotypic characteristics were not unique to any one genetic subpopulation (Chapter 3), the local genetic population subdivision (Chapter 2) suggests that this lake may possess unique combinations of quantitative traits, which can diversify the genome and raise the likelihood for long term persistence.  This population structure may be recently derived and in the process of increasing the level of subdivision or be a remnant of multiple distinctive populations colonizing the lake and slowly becoming homogenized. Although information from my thesis can not differentiate between these alternative hypotheses, I have been able to demonstrate the high level of diversity present both genetically and morphologically within the system along   77 with the presence of unique genetic characteristics that probably originated from different  glacial refugia, perhaps increase the lake trout’s adaptive potential. Lake trout have been called ‘glacial relicts’ due to their deep, cold water habitats and conceivably are better adapted to the past than the present (Wilson and Mandrak 2003).  The presence of glaciers at the south end of Atlin Lake appear to provide an ideal habitat for lake trout to thrive, whether it is because of the ‘proglacial lake like’ habitat or in spite of it is a question worth exploring in the future. The evolutionary potential of lake trout in this ecosystem depends on the survival of many unique subpopulations and conservation of the species requires protection of both intra and interspecific diversity (Ryman et al. 1995).  The diversity of lake trout present, both genetic and morphological, within this system may be a source to help restore diminished populations elsewhere.   Habitats similar to those in Atlin Lake may benefit from being seeded with populations adapted to such conditions.   Although a great many questions remain, my thesis has demonstrated that the Atlin Lake system possesses substantial intraspecific diversity. The level of divergence in this unperturbed system provides a snapshot on the time course of evolution in this long lived species and has presented a unique perspective on the evolutionary history of this wide-ranging species of northern Canada.       78 References  Adams, C.E., Woltering, C. Alexander, G. 2003. Epigenetic regulation of trophic morphology through feeding behaviour in Arctic charr, Salvelinus alpinus. Biol. J. Linn. Soc. 78:43-49  Allendorf, F.W., Leary, R.F. 1988. Conservation and distribution of genetic variation in a polytypic species, the cutthroat trout. Conservation Biology. 2:170-184  Alfonso, N.R. 2004. Evidence for two morphotypes of lake charr, Salvelinus namaycush, from Great Bear Lake, Northwest Territories, Canada. Environmental Biology of Fishes. 7:(11)21-23 Amos, W., Balmford, A.  2001. When does conservation genetics matter? Heredity. 87:257-265 Angers, B., Bernatchez, L. 1996. Usefulness of heterologous microsatellites obtained from brook charr, Salvelinus forntinalis Mitchill, in other Salvelinus species. Molecular Ecology. 5:317-319  Angers, B., Bernatchez, L. 1998. Combined use if SMM and non-SMM methods to infer fine structure and evolutionary history of closely related brook charr (Salvelinus fontinalis, Salmonidae) populations from microsatellites. Molecular Biological Evolution. 15(2):143-159  Angers, B., Bernatchez, L., Angers, A., Desgroseillers, L. 1995. Specific microsatellite loci for brook charr reveal strong population subdivision on a microgeographic scale. Journal of Fish Biology. 47(Suppl. A):177-185  Atlin Community Fisheries Working Group. 2001. Spawning report. unpub.  Avise, J.C., Hamrick J.L. eds. 1996. Conservation genetics: case histories from nature. Chapman and Hall, New York.  Avise, J.C., Helfman, G.S. Saunders, N.C., Hales, L.S. 1986. Mitochondrial DNA differentiation in North Atlantic eels: population genetic consequences of an unusual life history pattern. Proc Natl Acad Sci USA. 83:4350-4354  Bartley, D., Bagley, M., Gall, G., Bentley, B. 1992. Use of linkage disequilibrium data to estimate effective size of hatchery and natural fish populations. Conservation Biology. 6(3):365-375  Baxter, J.S., Taylor, E.B., Devlin, R.H., Hagen, J., McPhail, J.D. 1997. Evidence for natural hybridization between Dolly Varden (Salvelinus malma) and   79 bull trout (Salvelinus confluentus) in a northcentral British Columbia watershed. Can. J. Fish. Aquat. Sci. 54:421-429  Behnke, R.J., 1972. The systematics of salmonid fishes of recently glaciated lakes. Journal of the Fisheries Research Board of Canada. 29: 639–671.  Belkhir, K., Borsa, P., Goudet, J., Chikhi, L., Bonhomme, F. 1996–98  genetix, logiciel sous WindowsTM pour la génétique des populations. Laboratoire Génome et Populations, CNRS UPR 9060, Université de Montpellier II, Montpellier (France)  Belkhir, K., Castric, V., Bonhomme, F.  2002  IDENTIX, a software to test for relatedness in a population using permutation methods. Molecular Ecology Notes. 2:611-614  Bernatchez, L., Gle´met, H., Wilson, C.C., Danzmann, R.G. 1995. Introgression and fixation of arctic char (Salvelinus alpinus) mitochondrial genome in an allopatric population of brook trout (Salvelinus fontinalis). Canadian Journal of Fisheries and Aquatic Sciences. 52:179–185.  Bernatchez, L. Rhydderch, J.G., Kircheis, F.W. 2002. Microsatellite gene diversity analysis in landlocked Arctic char from Maine.  Transactions of American Fisheries.  131: 1106-1118  Blackie, C.T., Weese, D.J., Noakes, D.L.G. 2003. Evidence for resource polymorphism in the lake charr (Salvenlinus namaycush) population of Great Bear Lake, Northwest Territories, Canada. Ecoscience. 10:509-514  Bowen, B.W., 1999. Preserving genes, species, or ecosystems? Healing the fractured foundations of conservation policy. Molecular Ecology. 8: S5- S10  Bronte, C., Holey, M.E., Magenjian, C.P., Jonas, J.L., Claramunt, R.M. 2007. Relative abundance, site fidelity, and survival of Adult Lake Trout in Lake Michigan from 1999 to 2001: implications for future restoration strategies. North American Journal of Fisheries Management. 27:137-155  Bronte, C.R., Moore, S.A. 2007. Morphological variation of siscowet lake trout in Lake Superior. Transactions of American Fisheries Society 136:509-517  Bruford, M.W., Wayne, R.K. 1993. Microsatellites and their application to population genetic studies. Current Opinion in Genetics and Development. 3: 939-943    80 Brown, E.H. Jr., Eck, G.W., Foster, N.R., Horrall, R.M., Coberley, C.E. 1981. Historical evidence for discrete stocks of lake trout (Salvelinus namaycush) in Lake Michigan. Canadian Journal of Fisheries and Aquatic Sciences. 38(12):1747-1758  Burnham-Curtis, M.K., Smith, G.R.1994. Osteological evidence of genetic divergence of lake trout (Salvelinus namaycush) in Lake Superior. Copeia. 4:843-850  Caughley, G. 1994. Directions in conservation biology. The Journal of Animal Ecology. 63(2):215-244  Choisy, M., Franck, P., Cornuet, J-M. 2004. Estimating admixture proportions with microsatellites: comparison of methods based on simulated data. Molecular Ecology. 13:955-968  Cornuet, J-M., Piry, S., Luikart, G., Estoup, A., and Solignac, M. 1999. New methods employing multilocus genotypes to select or exclude populations as origins of individuals. Genetics. 153:1889-200  Costello, A.B., Down, T.E., Pollard, S.M., Pacas, C.J., and Taylor, E.B. 2003. The influence of history and contemporary stream hydrology on the evolution of genetic diversity within species: and examination of microsatellite DNA variation in bull trout, Salvelinus confluentus, (Pisces: Salmonidae). Evolution. 57: 328-344.  Coyne, J.A, Orr, H.A. 2004. Sympatric Speciation. In: Speciation.  Sinauer Associates, Inc. Sunderland, Massachusetts, USA  Crane, P.A., Lewis, C.J., Kretochmer, E.J., Miller, S.J., Spearman, W.J., DeCicco, A.L., Lisac, M.J., Weaburg, J.K. 2004. Characterization and inheritance of seven microsatellite loci from Dolly Varden, Salvelinus malma, and cross species amplification in Arctic char, S. alpinus. Con. Genetics. 5:737-741  David, J.R., Capy, P. 1988. Genetic variation of Drosphila melanogaster natural populations. Trends in Genetics 4:106-1111  DeHaan,P.W., Ardren,W.R. 2005. Characterization of 20 highly variable tetranucleotide microsatellite loci for bull trout (Salvelinus confluentus) and cross-amplification in other Salvelinus species. Molecular Ecology Notes. 5(3):582-585   Dehring, T.R., Brown, A.F., Daugherty, CH., Phelps, SR. 1981. Survey of the genetic variation among eastern lake Superior Lake trout (Salvelinus   81 namaycush). Canadian Journal of Fisheries and Aquatic Sciences. 38:1738-1746  DeKoning, J., Keatley, K., Phillips, R., Rhydderch, J., Janssen, J., Noakes, M. 2006. Genetic analysis of wild lake trout embryos recovered from Lake Michigan. Transactions of the American Fisheries Society. 135:399-407  DeWoody, J.A., Avise, J.C. 2000. Microsatellite variation in marine, freshwater and anadromous fishes compared with other animals.  Journal of Fish Biology. 56:461-473  Endler, J.A. 1987. Predation, light intensity and courtship behaviour in Poecila reticulate (Pisces: Poecilidae). Animal Behaviour. 35(5):1376-1385  Eshenroder, R.L., Krueger, C.C.  2002. Reintroduction of native fishes to the Great Lakes proper: a research theme area. Great Lakes Fishery Commission  Evanno, G., Regnaut, S., Goudet, J.  2006. Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Molecular Ecology. 14:2611-2620  Excoffier, L., Smouse, P., and Quattro, J.M. 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics. 131:479-491  Excoffier, L., Laval, G. ,Schneider, S. 2005. Arlequin ver 3.0: An integrated software package for populations genetics data analysis. Evolutionary Bioinformatics Online. 1:47-50  Fraley, D., Raftery, A.E. 2003 Enhanced software for model based clustering, discriminant analysis and density estimation: MCLUST. Journal of Classification. 20:263-286. http://www.stat.washington.edu/fraley/mclust  Frankham, R. 1995. Conservation genetics. Ann. Rev. Genet. 29:305-327  Frankham, R., Briscoe, D.A., Ballou, J.D. 2002. Introduction to Conservation Genetics. Cambridge University Press, New York, New York, USA.  Forbes, S.H., Hogg, J.T., Buchanan, F.C., Crawford, A.M., Allendorf, F.W. 1995. Microsatellite evolution in congeneric mammals: domestic and bighorn sheep.  Molecular Biology and Evolution. 12:1106-1113  Gaggiotti, O.E., Lange, O., Rassmann, L., and Gliddon, C. 1999. A comparison of two indirect methods for estimating average levels of gene flow using microsatellite data. Molecular Ecology. 8: 513-520.   82  Grewe, P.M., Krueger, C.C., Marsden, J.E., Aquadro, C.F., May B. 1994. Hatchery origins of naturally Produced produced lake trout fry captured in Lake Ontario: temporal and spatial variability based on allozyme and mitochondrial DNA data. Transactions of American Fisheries. 123:309-320  Griffiths, D. 1997. Local and regional species richness in North American lacustrine fish. The Journal of Animal Ecology.  66(1):49-56  Goudet ,J. 1999.  PCA-GEN for Windows.  Available from the Institute of Ecology, Biology Building. University of Lausanne, Switzerland.  Goudet ,J. 2001.  FSTAT version 2.9.3.1 Updated from Goudet J. 1995. J. Heredity. 86:485-486  Goudet, J., Raymond, M., DeMeeüs, T., Rousset, F. 1996. Testing differentiation in diploid populations. Genetics. 144:933-1940  Gow, J.L., Peichel, C.L., Taylor, E.B. 2007. Ecological selection against hybrids in natural populations of sympatric threespine sticklebacks. Journal of Evolutionary Biology. 20:2173-2180  Guinand, B., Scribner, K.T. 2003. Evaluation of methodology for detection of genetic bottlenecks: inference from temporally replicated lake trout populations. Comptes rendus Biologies 36(HS1):61-67  Guinand, B., Scribner, K.T., Page, K.S., Burnham-Curtis, MK. 2003. Genetic variation over space and time: analyses of extinct and remnant lake trout populations in the Upper Great Lakes. Proc. R. Soc. Lond., Ser. B: Biological Sciences. 270(1513): 425-433  Gunn, J.M., Steedman, R.J., Ryder, R.A. 2003.  Boreal Shield Watersheds: Lake Trout Ecosystems in a Changing Environment (Chapters 1-3). CRC Press: Boca Raton, FL.  Hammer, Ø., Harper, D.A.T., Ryan., P.D. 2001. PAST: Palaeontological Statistics software package for education and data analysis. Palaeontological Electronica 4(1): 9  Hanson, J.M., Leggett, W.C. 1982. Empirical predications of fish biomass and yield. Canadian Journal of Fisheries and Aquatic Sciences. 39: 257-263  Hedrick, P.W. 2001. Conservation genetics: where are we now? Trends in Ecology & Evolution. 16(11): 629-636   83  Henderson, B.A., Anderson, D.M. 2002. Phenotypic differences in buoyancy and energetics of lean and siscowet lake charr in Lake Superior. Environmental Biology of Fishes. 64:203-209  Hendry, A.P., Stearns, S.C. 2004. Evolution Illuminated: Salmon and Their Relatives. Oxford University Press, New York, N.Y.  Hendry, A.P., Taylor, E.B. 2004. How much of the variation in adaptive divergence can be explained by gene flow?  An evaluation using lake- stream stickleback pairs. Evolution. 58(10):2319-2331  Hewitt, G. 1996. Some genetic consequences of ice ages, and their role in divergence and speciation. Biological Journal of the Linnean Society. 58:247-276  Hilborn, R. 2007. Managing fisheries is managing people: what has been learned? Fish and Fisheries 8:285–296.  Ihssen, P.E., Casselman, J.M, Martin, G.W, Phillips, R.B.  1988. Biochemical genetic differentiation of lake trout (Salvelinus namaycush) stocks of the Great Lakes region. Canadian Journal of Fisheries and Aquatic Sciences. 45: 1015-1029  Imre, I., McLauglin, R.L., Noakes, D.L.G. 2002. Phenotypic plasticity in brook charr: changes in caudal fin induced by water flow. Journal of Fish Biology. 61:1171-1181  Kalinowski, S.T. 2003. Genetic mixture analysis 1.0. Department of Ecology, Montana State University, Bozeman, Mont., USA. Available from http://www.montana.edu/kalinowski/GMA/GMA_Home.html  Keeley, E.R., Parkinson, E.A., and E.B. Taylor. 2007. The origins of ecotypic variation of rainbow trout: a test of environmental vs. genetically based differences in morphology. Journal of Evolutionary Biology. 20: 725–736.  Kirkland, R.A., Gray, C.B.J. 1986. Reconnaissance of the Chemical and Biological Limnology in Four Large Lakes of the Yukon River Basin. Inland Waters Directorate, National Hydrology Research Institute, Saskatoon, Saskatchewan.  Krueger, C.C., Ihssen, P.E.1995. Review of genetics of lake trout in the Great Lakes: History, molecular genetics, physiology, strain comparisons, and restoration management. Journal of Great Lakes Research. 21(suppl. 1): 348-363    84 Kruger, C.C., Marsden, J.E., Kincaid, H., May, B. 1989. Genetic differentiation among lake trout strains stocked into Lake Ontario. Transactions of the American Fisheries Society. 118:317-330  Larkin, P.A. 1981. A perspective on population genetics and salmon management. Canadian Journal of Fisheries and Aquatic Sciences. 38:1469-1475  Larsen, P. F., Hansen, M.M., Nielsen, E.E., Jensen, L.F., Loeschcke, V. 2005. Stocking impact and temporal stability of genetic composition in a brackish northern pike population (Esox lucius L.), assessed using microsatellite DNA analysis of historical and contemporary samples. Heredity. 95:136– 143.  Latch, E.K., Dharmarajan, G., Glaubitz, J.C., Rhodes, O. E. Jr. 2006. Relative performance of Bayesian clustering software for inferring population substructure and individual assignment at low levels of population differentiation. Conservation Genetics. 7:295-302   Latta, R.G. 1998. Differentiation of allelic frequencies at quantative trait loci affecting locally adaptive traits. The American Naturalist. 151(3): 283-292  Li,Y.-C., Korol, A.B., Fahima, T., Beiles, A., Nevo, E. 2002. Microsatellites: genomic distribution, putative function and mutational mechanisms: a review. Molecular Ecology. 11:2453-2465  Lindsay, C.C., McPhail, J.D. 1986. Zoogeography of Fishes of the Yukon and Mackenzie Basins. In: The Zoogeography of North American Freshwater Fishes. (eds. Hocutt C.H., Wiley E.O.) John Wiley and Sons, New York. pp 639-674  Lindsay, C.C., Patalas, K., Bodaly, R.A., Archibald, C.P. 1981. Glaciation and the Physical, Chemical and Biological Limnology of Yukon Lakes. Technical Report of Fisheries and Aquatic Sciences, Winnipeg, Man. pp 45 Lynch, M. 1997. Inbreeding and outbreeding depression. Pg 59-70 in Genetic Effects of Straying of Non-Native Hatchery Fish into Natural Populations, W.S.Grant ed. NOAA Tech. Memo NMFS-NWFSC-30 Magnan, P., Audet, C., Glemet, H., Legault, M., Rodriguez, A., Taylor, E.B. 2002. Developments in the ecology, evolution, and behaviour of the charrs, genus Salvelinus: relevance for their management and conservation. Environmental Biology of Fishes. 64:9-14    85 Mantel, N. 1969. The detection of disease clustering and a generalized regression approach. Cancer Research. 27:209-220  Mathieu, E., Autern, M., Roux, M., Bonhomme, F. 1990. Épreves de validation dans l’analyze de structures génétiques multivariées`comment tester l’équilibre panmictique? Revue de Statistique Appliquée. 38:47-66  Miller, M.P. 1997. Tools for population genetic analysis (TFPGA) 1.3: A Windows program for the analysis of allozyme and molecular population genetic data. Computer software distributed by author.  Moore, S.A., Bronte, C.R. 2001. Delineation of sympatric morphotypes of lake trout in Lake Superior. Transactions of the American Fisheries. 130:1233- 1240  Morris, D.B., Richard, K., Wright, J.M.1996. Microsatellites from rainbow trout (Oocorhynchus mykiss) and their use for genetic study of salmonids. Canadian Journal of Fisheries and Aquatic Sciences. 53:120-126  Nei, M., Maruyama, T., Chakraborty, R. 1975. The bottleneck effect and genetic variation in populations. Evolution. 29:1-10  Nei, M., Tajima, F. 1981. Genetic drift and estimation of effective population size. Genetics. 98:625-640  Neilsen, J.L., Gan, C.A., Wright, J.M., Thomas, K. 1994.  Biogeographic distributions of mitochondrial and nuclear markers for southern steelhead. Molecular Marine Biology and Biotechnology. 3:281-293  Neilsen, E.E., Hansen, M.M., Loeshcke, V. 1999. Genetic variation in time and space: microsatellite analysis of extinct and extant populations of Atlantic Salmon. Evolution. 53(1):261-268  Ohta, T., Kimura, M. 1973. A model of mutation appropriate to estimate the number of electrophoretically detectable alleles in a finite population. Genet. Res. 22:201–204.  O’Reilly, P.T., Hamilton, L.C., McConnell, S.K., Wright, J.M. 1996. Rapid analysis of genetic variation on Atlantic salmon (Salmo salar) by PCR multiplexing of dinucleotide and tetranucleotide microsatellite. Canadian Journal of Fisheries and Aquatic Sciences. 53:2292-2298  Østbye, K., Amundsen, P.A., Bernatchez, L., Klemetsen, A., Knudsen, R., Kristoffersen, R., Næsje, T.F., Hindar, H. 2006. Parallel evolution of ecomorphological traits in the European whitefish Coregonus lavaretus   86 (L.) species complex during postglacial times. Molecular Ecology. 15(13):3983-4001  Paetkau, D., Strobeck, C. 1994. Microsatellite analysis of genetic variation in black bear populations. Molecular Ecology. 3:489-495  Page, K.S., Scribner, K.T., Bennett, K.R., Garzel, L.M., Burnham-Curtis, M.K. 2003. Genetic assessment of strain-specific sources of lake trout recruitment in the Great Lakes. Transactions of the American Fisheries Society. 13(25):877-894  Pakkasmaa, S., Piironen, J. 2001. Water velocity shapes juvenile salmonids, Evolutionary Ecology. 14:721-730  Peel, D., Overdon, J.R., Peel, S.L. 2004. NeEstimator: software for estimating effective population size, Version 1.3. Queensland Government, Department of Primary Industries and Fisheries.  Pella, J.J., Milner, G.B. 1987. Use of genetic marks in stock composition analysis. In: Population Genetics and Fishery Management (eds. Ryman and Utter) University of Washington Press. Seattle, WA.  pp 247-276  Perkins, D.L., Fitzsimons, J.D., Marsden, J.E., Krueger, C.C., May, B. 1995. Differences in reproduction among hatchery strains of lake trout at eight spawning areas in Lake Ontario: Genetic evidence from mixed stock analysis. Journal of Great Lakes Research. 21(suppl.1):365-374  Petit, R.J., El Mousadik, A., Pons, O. 1998. Identifying populations for conservation on the basis of genetic markers. Conservation Biology. 12(4):844-855  Phelps. R.F. and Knudsen, K.L. 1987. Genetic divergence and identification of seven cutthroat trout subspecies and rainbow trout. Transactions of American Fisheries. 116: 580-587  Piller, K., Wilson, C.C., Lee, C.E., Lyons, J. 2005. Conservation genetics of inland lake trout in the upper Mississippi River Basin: stocked or native ancestry? Transactions of the American Fisheries Society. 134:789-802  Piry, S., Alapetite, A., Cornuet, J-M., Paetkau, D., Baudouin, L., and Estoup, A. 2004. GENECLASS2: a software for genetic assignment and first generation migrant detection. Journal of Heredity. 95:536-539  Power, G. 2002. Charrs, glaciations and seasonal ice. Environmental Biology of Fishes. 64:17-35    87 Primmer, C. R. 2005. Genetic characterisation of populations and its use in conservation decision making in fish. The role of biotechnology. Turin, Italy. 3-5-2005. pp 137-142.  Pritchard J.K., Stephens, M., Donnelly, P. 2000. Inference of population structure using multilocus genotype data. Genetics. 155: 945-959  Rahal, F.J. 2002. Homogenization of freshwater faunas. Annu. Rev. Ecol. Syst. 33:291-315  Rannala, B., Mountain, J.L.1997. Detecting immigration by using multilocus genotypes. Proc. Natl. Aca. Sci. 94:9197-9201  Raybould, A.F., Mogg, R.J., Clarke, R.T., Gliddon, C.J., Gray, A.J. 1999. Variation and population structure at microsatellite and isozyme loci in wild cabbage (Brassica oleracea L.) in Dorset (UK). Genetic Resources and Crop Evolution. 46(4): 351-360  Raymond, M., F. Rousset, 1995. Genepop (version 1.2), population genetics software for exact tests and ecumenicism. Journal of Heredity 86: 248-249  Rice, W.R. 1989. Analyzing tables of statistical tests. Evolution. 43(1): 223-225  Rohlf, F.J. 1997. TPSDIG and TPSRELW programs. Available from http://life.bio.sunysb.edu/morph  Ryman, N. 1983. Patterns of distribution of biochemical genetic variation in salmonids: differences between species. Aquaculture. 33:1-21  Ryman, N., Utter, F. eds. 1987. Population Genetics and Fishery Management. University of Washington Press. Seattle, WA  Schluter, D., Clifford, E.A., Nemethy, M., McKinnon, J.S. 2004. Parallel evolution and inheritance of quantitative traits. The American Naturalist. 163:809- 822 Schwartz, M.K., Tallmon, D.A., Luikart, G. 1998. Review of DNA-based census and effective population size estimators. Animal Conservation.  2:320-322  Scott, W. B. and Crossman, E. J. 1973. Freshwater fishes of Canada. Bull. Fish. Res. Board Can. pp 220-227  Slatkin, M. 1985. Gene flow in natural populations. Annu. Rev. Ecol. Syst. 16:393-430  Slatkin, M. 1993. Isolation by distance in equilibrium and non-equilibrium populations. Evolution. 47(1):264-279   88  Slatkin, M. 1987. Gene flow and the geographic structure of natural populations. Science. 236:787-792  Shortreed, K.S., Stockner, J.G. 1983. Limnology of Selected Lakes in the Yukon River Basin. DFO, West Vancouver, BC. Yukon River Basin Study  Slatkin, M. 1985. Rare alleles as indicators of gene flow. Evolution. 39 (1): 53-65  Soulé, M. 1973. The epistasis cycle: a theory of marginal populations. Annu. Rev. Ecol. Syst. 4:165-187 Srivastava, D.S. 2002. The role of conservation in expanding biodiveristy research. Oikos. 98(2):351-360 Sunnucks, P. 2000. Efficient genetic markers for population biology. Trends in Ecology and Evolution. 15(5): 199-203.  Tamkee, P. 2006. The influence of history and contemporary stream hydrology on the evolution of genetic diversity within species: an examination of microsatellite DNA variation in rainbow trout (Oncorhynchus mykiss). MSc Thesis, University of British Columbia, Vancouver, BC.  Tamkee, P., Taylor, E.B. 2003. Genetic analysis of lake trout (Salvelinus namaycush) collected from Banff National Park, AB  Taylor, E.B. 1999. Species pairs of north temperate freshwater fishes: evolution, taxonomy and conservation. Reviews in Fish Biology and Fisheries. 9:299-334  Taylor, E.B., Costello, A.B. 2006. Microsatellite DNA analysis of coastal populations of bull trout (Salvelunus confluentus) in British Columbia: zoogeographic implications and its application to recreational fishery management. Canadian Journal of Fisheries and Aquatic Sciences. 63:1157-1171  Taylor, E.B., Stamford, M.D., Baxter, J.S. 2003. Population subdivision in westslope cutthroat trout (Oncorhynchus clarki lewisi) at the northern perphery of its range: evolutionary inferences and conservation implications. Molecular Ecology.122:609-622  Vitic, R., Strobeck, C. 1996. Mitochondrial DNA analysis of populations of lake trout (Salvelinus namaycush) in west-central Canada: implications for stock identification. Canadian Journal of Fisheries and Aquatic Sciences. 52:1038-1047)    89 Volis, S., Yakubov, B., Shulgina, I., Ward, D., Mendlinger, S. 2005. Distinguishing adaptive from nonadaptive genetic differentiation: comparison of QST and FST at two spatial scales. Heredity. 95:466-475  Waples, R.S., Gaggiotti, O. 2006. What is a population? An empirical evaluation of some genetic methods for identifying the number of gene pools and their genetic connectivity. Molecular Ecology. 15: 1419-1439  Weir, B.S. and Cockerham, C.C. 1994. Estimating F-statistics for the analysis of population structure. Evolution. 38(6):1358-1370  Wenburg, J.K., Benzen, P., Foote, C.J. 1998. Microsatellite analysis of population structure in an endangered salmonid: the coastal cutthroat trout (Oncorhynchus clarki clarki). Molecular Ecology. 7:733-750  Westrich, K.M., Konkol, N.R., Matsuoka, M.P., Phillips, R.B. 2002. Interspecfic relationships among charrs based on phylogenetic analysis of nuclear growth hormone intron sequences. Environmental Biology of Fishes. 64:217-222  Whitlock, M.C. 2003. Fixation probability and time in subdivided populations. Genetics. 164(2):767-779  Wilher, I.L.1956. A Limnological Survey of Atlin and Southern Tagish Lakes. Management Publication No. 5 of the British Colombia Commission.  Wilson, C.C, Bernatchez, L. 1998 The ghost of hybrids past: Fixation of arctic charr (Salvelinus alpinus) mitchondrial DNA in an introgressed population of lake trout (S. namaycush). Molecular Ecology. 71:127-132  Wilson, C.C., and Hebert, P.D.N. 1993. Natural hybridization and introgression between lake trout (Salvelinus namaycush) and arctic charr (S. alpinus) in the Canadian arctic. Canadian Journal of Fisheries and Aquatic Sciences. 50: 2652-2658.  Wilson, C.C., Hebert, P.D.N. 1996. Phylogeographic origins of lake trout (Salvelinus namaycush) in eastern North America. Canadian Journal of Fisheries and Aquatic Sciences. 53: 2764-2775  Wilson, C.C., Hebert, P.D.N. 1998. Phylogeography and postglacial dispersal of lake trout (Salvelinus namaycush) in North America. Canadian Journal of Fisheries and Aquatic Sciences. 55:1010-1024  Wilson, C.C., Mandrak, N.E. 2003. History and evolution of lake trout in shield lakes: past and future challenges In: Boreal Shield Watersheds: Lake   90 Trout Ecosystems in a Changing Environment (eds Gunn, Steedman, Ryder). CRC Press: Boca Raton, FL. pp 21-33  Zamudio, K.R., Wiecozorek, A.M. 2007. Fine-scale spatial genetic structure and dispersal among spotted salamander (Ambystoma maculatum) breeding populations. Molecular Ecology. 16:257-274  Zimmerman, M.S., Krueger, C.C., Eshenroder, R.L. 2006. Phenotypic diversity of lake trout in Great Slave Lake: Differences in morphology, buoyancy, and habitat depth. Transactions of the American Fisheries Society. 135: 1056- 1067  Zimmerman, M.S., Krueger, C.C., Eshenroder, R.L. 2007. Morphological and ecological differences between shallow and deep-water lake trout in Lake Mistassini, Quebec.  Journal of Great Lakes Research. 33:156-169   http://earth.google.com/download-earth.html    91 Table 1: Sample site names, number of samples, locations, watersheds, elevations and surface area Lake # Lake  Location Sample Size  Latitude Longitude Watershed Elevation (m) Area (km2) 1 Atlin (AL) BC/Yukon 183 59o 29'57.14"N 133o 44'12.07"W Yukon River 886 634.5 2 Tatsemerie  (T) British Columbia 35 58o 20'47.47"N 132o 19'14.49"W Taku River 1270 ~57 3 Tagish  (TAG) BC/Yukon 227 59o 46'21.87"N 134o 14'29.46"W Yukon River 1109 354.6 4 Trapper  (TR) British Columbia 9 58o 27' 27"N 132o 37' 32W Taku River 1457 ~24 5 Muncho  (M) British Columbia 6 58o 59' 38"N 125o 46' 57"W Peace River 1494 ~50 6 Frances (F) Yukon 19 61o 23' N 129o 35' W Mackenzie River 804 99.41 7 Traviallant (TV) NWT 17 69o 39'16.95N 131o 5l'03.75"W Mackenzie River 27 ~102 8 Bednesti (B) British Columbia 25 53o 22' 41"N 123o 51' 44"W Fraser River 1201 2.61 9 Hautete (H) British Columbia 32 55o 11' 14"N 126o 8' 50"W Fraser River 989 2.3 10 Tezzeron (TE) British Columbia 23 54o 42' 52"N 124o 27' 59W Fraser River 910 78.08 11 Arctic (AR) British Columbia 8 54o 25' 17"N 121o 40' 36"W Fraser River 1327 1.09 12 Cluculz (CL) British Columbia 24 53o 52' 23"N 123o 34' 20"W Fraser River 860 19.88 13 Fraser (FRA) British Columbia 27 55o 4' 56"N 124o 45' 3"W Fraser River 880 54.6 14 Nakinlerak (K) British Columbia 36 55o 14' 52.512"N 125o 14' 54.134"W Fraser River 1092 7.34 15 Natowite (N) British Columbia 32 55o 5' 11"N 125o 56' 46"W Fraser River 938 16 16 Cunningham (CU) British Columbia 30 54o 35' 31"N 125o 19' 20"W Fraser River 863 29.96 17 Airline (A) British Columbia 28 55o 7' 43"N 125o 3' 47"W Fraser River 839 4.46 18 Klawli (KL) British Columbia 28 55o 21' 0"N 124o 21' 43"W Fraser River 1006 4.58 19 Pinchi (P) British Columbia 26 54o 36' 43"N 124o 24' 28"W Fraser River 830 55.54     92 Table 2: Locus names, size range of alleles in base pairs, annealing temperatures, total numbers of alleles (N), and references, for microsatellite DNA loci assayed in western lake trout used in the current study.  Name Size Annealing Temp. (oC) N Source Species Source Smm22 132-238 55 30 Dolly Varden Crane et al. 2004 Sco102 144-168 55 4 Bull Trout Bettles, Jones, Young unpublished1 Ssa197 199-307 58/56 24 Atlantic Salmon O'Reilly et al. 1996 Sco19 144-182 48/47 15 Bull Trout E.B. Taylor unpublished Sfo18 162-188 56/55 11 Brook Trout Angers et al. 1995 Sco107 200-272 55 16 Bull Trout Bettles, Jones, Young unpublished1 Sco215 283-319 55 9 Bull Trout DeHaan and Arden 2005 Sco2 161-207 55 13 Bull Trout E.B. Taylor unpublished  1Washington Department of Fish and Game, Olympia, WA     93 Table 3: Summary of allelic variation at eight microsatellite loci in lake trout. Number of alleles per locus (A), expected heterozygosity (He), observed heterozygosity (Ho), allelic richness (Ar ), and number of individuals genotyped (N) for each loci per population.  Sample locations are defined in Table 1.  Location SMM22 SCO102 SSA197 SCO19 SFO18 SCO107 SCO215 SCO2 Average AL A 26 3 18 11 7 10 8 6 11.125 He 0.941 0.3028 0.8789 0.8123 0.8167 0.818 0.5436 0.8305 0.7430 Ho 0.913 0.3026 0.8324 0.7377 0.6962 0.711 0.5 0.4321 0.6406 Ar 3.676 1.605 3.357 3.04 3.058 3.065 2.18 2.142 2.7654 N 184 185 179 183 158 173 170 162 174.25 T A 16 4 9 6 4 6 5 7 7.125 He 0.8694 0.6691 0.8473 0.6475 0.534 0.7224 0.5874 0.7784 0.7069 Ho 0.7714 0.6791 0.8571 0.6875 0.4231 0.7429 0.5517 0.72 0.6791 Ar 3.375 2.502 3.239 2.433 2.019 2.705 2.232 2.938 2.6804 N 35 34 35 32 26 35 29 25 31.375 TAG A 23 4 19 10 7 11 6 10 11.25 He 0.9069 0.2964 0.8786 0.7433 0.756 0.825 0.5079 0.6223 0.6921 Ho 0.8678 0.2788 0.837 0.6607 0.6108 0.795 0.5088 0.7195 0.6598 Ar 3.506 1.596 3.35 2.779 2.824 3.108 2.071 2.361 2.6994 N 227 226 227 224 203 225 226 221 222.375 TR A 7 3 4 3 2 4 2 1 3.25 He 0.8148 0.4753 0.5926 0.4259 0.1975 0.5741 0.4444 0 0.4406 Ho 1 0.3333 0.8889 0.451 0 0.7778 0.6667 0 0.5147 Ar 3.248 1.984 2.295 1.893 1.405 2.26 1.833 1 1.9898 N 9 9 9 9 9 9 9 9 9 M A 5 1 3 6 2 3 4 1 3.125 He 0.7222 0 0.5417 0.8056 0.4444 0.2917 0.7083 0 0.4392 Ho 1 0 0.3333 0.8788 0 0.1667 0.8333 0 0.4015 Ar 2.917 1 2.182 3.309 1.933 1.667 2.826 1 2.1043 N 6 5 6 6 3 6 6 5 5.375 F A 13 2 9 6 3 5 3 4 5.625 He 0.5647 0.4654 0.8488 0.7618 0.6172 0.644 0.59 0.5833 0.6344 Ho 0.7895 0.6316 0.8889 0.6316 0.5625 0.7895 0.4118 0.3889 0.6368 Ar 3.542 1.842 3.308 2.9 2.33 2.442 2.229 2.29 2.6104 N 19 19 18 19 16 19 17 18 18.125 TV A 9 1 5 4 4 7 5 5 5 He 0.5277 0 0.7526 0.6403 0.199 0.7561 0.6869 0.7289 0.5364 Ho 0.5882 0 0.5882 0.5857 0.1429 0.7059 0.2941 0.3333 0.4048 Ar 2.22 1 2.82 2.484 1.429 2.857 2.601 2.836 2.2809 N 17 17 17 14 14 17 17 15 16    94 Table 3 continued…   SMM22 SCO102 SSA197 SCO19 SFO18 SCO107 SCO215 SCO2 Average B A 7 1 7 2 1 3 1 6 3.5 He 0.7224 0 0.6312 0.32 0 0.0799 0 0.6184 0.2965 Ho 0.68 0 0.68 0.08 0 0.0833 0 0.68 0.2754 Ar 2.802 1 2.555 1.661 1 1.31 1 2.48 1.7260 N 25 25 25 25 24 24 24 25 24.625 H A 7 1 11 5 3 5 4 3 4.875 He 0.6104 0 0.8408 0.3706 0.3261 0.6177 0.5604 0.6183 0.4930 Ho 0.6563 0 0.7188 0.375 0.3333 0.6563 0.5161 0.6292 0.4856 Ar 2.423 1 3.23 1.784 1.631 2.391 2.189 2.303 2.1189 N 32 30 32 32 30 32 31 29 31 TE A 8 1 12 9 2 6 3 4 5.625 He 0.6938 0 0.8866 0.7098 0.3967 0.6446 0.5983 0.6043 0.5668 Ho 0.8261 0 0.913 0.6957 0.2727 0.6522 0.5217 0.4286 0.5388 Ar 2.643 1 3.479 2.698 1.731 2.423 2.274 2.304 2.3190 N 23 23 23 23 22 23 23 21 22.625 AR A 4 2 6 5 1 1 3 4 3.25 He 0.5391 0.2449 0.6327 0.625 0 0 0.5547 0.4844 0.3851 Ho 0.625 0 0.7143 0.5 0 0 0.5917 0.375 0.3508 Ar 2.22 1.505 2.633 2.538 1 1 2.142 2.104 1.8928 N 8 7 7 8 2 7 8 8 6.875 C A 3 1 10 5 4 4 1 4 4 He 0.4688 0 0.8446 0.3854 0.5103 0.4244 0 0.5076 0.3926 Ho 0.5 0 0.9167 0.4583 0.5455 0.5217 0 0.7826 0.4656 Ar 1.914 1 3.257 1.824 2.089 1.903 1 1.987 1.8718 N 24 24 24 24 22 23 24 23 23.5 FRA A 5 1 12 8 3 6 3 4 5.25 He 0.6852 0 0.8717 0.6303 0.4907 0.5473 0.5652 0.513 0.5379 Ho 0.5556 0 0.963 0.7407 0.5 0.5185 0.4444 0.4815 0.5255 Ar 2.57 1 3.387 2.489 1.878 2.134 2.194 2.039 2.2114 N 27 27 27 27 22 27 27 27 26.375 K A 5 1 12 6 2 3 5 3 4.625 He 0.3557 0 0.8229 0.7014 0.4348 0.434 0.62 0.6269 0.4995 Ho 0.2778 0 0.9167 0.8611 0.5833 0.4444 0.5833 0.9444 0.5764 Ar 1.747 1 3.157 2.638 1.785 1.846 2.324 2.319 2.1020 N 36 36 36 36 36 36 36 36 36 N A 10 1 12 6 4 9 3 4 6.125 He 0.8081 0 0.8618 0.6483 0.2051 0.7158 0.5327 0.646 0.5522 Ho 0.6875 0 0.8755 0.6129 0.2089 0.8438 0.5313 0.5625 0.5403 Ar 3.084 1 3.333 2.49 1.428 2.705 2.046 2.389 2.3094 N 32 32 32 31 27 32 32 32 31.25   95 Table 3 continued…   SMM22 SCO102 SSA197 SCO19 SFO18 SCO107 SCO215 SCO2 Average CU A 11 1 17 4 2 3 2 5 5.625 He 0.8528 0 0.9172 0.515 0.4061 0.4961 0.095 0.6956 0.4972 Ho 0.9333 0 0.9667 0.4667 0.5 0.5 0.0333 0.5517 0.4940 Ar 3.28 1 3.618 2.084 1.742 1.967 1.19 2.581 2.1828 N 30 30 30 30 30 30 30 29 29.875 A A 4 1 8 3 2 2 1 4 3.125 He 0.4917 0 0.7953 0.203 0.4032 0.1327 0 0.4767 0.3128 Ho 0.5714 0 0.9286 0.2222 0.32 0.0714 0 0.2963 0.3012 Ar 2.045 1 3.022 1.416 1.74 1.263 1 1.948 1.6793 N 28 28 28 27 25 28 28 27 27.375 KL A 7 1 5 4 4 2 2 3 3.5 He 0.7449 0.6958 0.4094 0.4094 0.2713 0.0351 0.4938 0.4904 0.4438 Ho 0.7778 0.6786 0.3929 0.3929 0.1304 0.0357 0.4444 0.3214 0.3968 Ar 2.822 1 2.609 1.803 1.569 1.071 1.88 1.909 1.8329 N 27 28 28 28 23 28 27 28 27.125 PL A 8 1 14 6 2 3 3 4 5.125 He 0.7078 0 0.8809 0.716 0.2392 0.5324 0.4408 0.733 0.5313 Ho 0.6538 0 1 0.6154 0.1667 0.5217 0.48 0.7692 0.5259 Ar 2.753 1 3.44 2.716 1.466 1.968 1.862 2.724 2.2411 N 26 26 26 26 18 23 25 26 24.5 GLOBAL A 30 4 24 15 7 16 9 13 14.75 He 0.6857 0.1658 0.7756 0.5827 0.3815 0.4890 0.4489 0.5557 0.5106 Ho 0.7197 0.1528 0.8006 0.5607 0.3156 0.5020 0.4165 0.4956 0.4954 Ar 3.657 1.494 3.453 3.276 2.709 3.215 2.759 2.605 2.896 N 815 811 809 804 710 797 789 766 787.625   96 Table 4: Estimates of pairwise FST values (mean over 8 loci) for 19 western lake trout lakes. Population codes are defined in Table 1.   AL T TAG TR M F TV B H TE AR C FRA K N CU A KL T 0.076 TAG 0.014 0.072 TR 0.236 0.200 0.237 M 0.195 0.223 0.200 0.440 F 0.064 0.067 0.082 0.277 0.189 TV 0.212 0.237 0.223 0.442 0.292 0.233 B 0.327 0.371 0.333 0.616 0.561 0.419 0.415 H 0.201 0.263 0.216 0.442 0.360 0.225 0.256 0.337 TE 0.193 0.232 0.202 0.424 0.310 0.218 0.196 0.294 0.072 AR 0.275 0.301 0.279 0.497 0.439 0.329 0.306 0.516 0.338 0.178 C 0.254 0.322 0.262 0.528 0.450 0.303 0.381 0.474 0.192 0.132 0.403 FRA 0.205 0.256 0.212 0.449 0.331 0.245 0.228 0.322 0.083 0.009 0.217 0.173 K 0.207 0.265 0.214 0.428 0.368 0.240 0.292 0.325 0.046 0.076 0.320 0.143 0.112 N 0.203 0.232 0.213 0.420 0.305 0.214 0.207 0.286 0.063 0.042 0.218 0.208 0.069 0.084 CU 0.218 0.286 0.228 0.465 0.363 0.250 0.276 0.409 0.174 0.075 0.277 0.155 0.125 0.179 0.147 A 0.317 0.386 0.325 0.612 0.559 0.411 0.332 0.464 0.336 0.241 0.388 0.501 0.263 0.371 0.299 0.344 KL 0.287 0.336 0.297 0.549 0.463 0.355 0.281 0.387 0.249 0.146 0.265 0.425 0.203 0.279 0.167 0.253 0.160 PL 0.196 0.251 0.214 0.438 0.334 0.219 0.224 0.288 0.065 0.012 0.254 0.151 0.054 0.094 0.075 0.066 0.258 0.167 * boldface values were significantly greater than 0.   97   Table 5: Log likelihood scores from STRUCTURE for all sample lakes. The boldface values represents most likely K based on the maximum value of ∆K.  K log (K) ∆K 1 -28121.9 N/A 2 -23931.3 3959.772 3 -23483.8 11.47574 4 -23089.9 0.61774 5 -22759.1 0.219316 6 -22406.2 0.923817 7 -22172.8 0.391189 8 -22006.6 0.797577 9 -21980.3 1.04857 10 -21789.5 0.459834 11 -21683.5 0.23141 12 -21620.6 0.149534 13 -21526.6 0.658318 14 -21554.6 1.265766 15 -21368.9 0.800788 16 -21347.6 0.492561 17 -21213.9 0.172127 18 -21118.8 0.137212 19 -20998.5 1.401446 20 -21091.9 N/A    98 Table 6: Analysis of molecular variance within and among western lake trout populations.     Variance Component Fstat Percent Variation P Fstat Percent Variation P Fstat Percent Variation P Between Regions 0.19582 19.58 0 0.14965 14.97 0 0.00422 0.42 0.08602 Among populations within regions 0.33338 13.76 0 0.13684 11.64 0 0.03599 3.58 0 Within populations among regions 0.17106 66.66 0 0.26602 73.4 0 0.04006 95.99    0 Yukon vs Taku vs Mackenzie vs Fraser Watersheds Among PCA Groups (19 Pops) Atlin subpopulations vs Tagish subpopulations   99 Table 7:  Estimates of effective population size of all sample lakes Lake # Lake  Sample Size  Ne 95% CI 1 Atlin  186 662.2 405.8 - 16124 2 Tatsemerie  35 31.7 25.4 - 40.8 3 Tagish  227 1845.8 823.7 - ∞ 4 Trapper  9 11.1 5.8 - 39.2 5 Muncho  6 3.1 1.9 - 5.8 6 Frances  19 21.7 14.7 - 36.9 7 Traviallant 17 60.1 23.6 - ∞ 8 Bednesti 25 61.9 19.6 - ∞ 9 Hautete 32 38.1 24.0 - 74.7 10 Tezzeron 23 111.8 42.2 - ∞ 11 Arctic 8 2.9 2.0 - 4.4 12 Cluculz 24 44.1 20.8 - 417.7 13 Fraser 27 121.5 44.7 - ∞ 14 Nakinlerak 36 70.1 37.3 - 267.4 15 Natowite 32 81.5 43.7 - 332.1 16 Cunningham 30 133.9 51.4 - ∞ 17 Airline 28 96.7 26.4 - ∞ 18 Klawli 28 232.8 239.2 - ∞ 19 Pinchi 26 19.5 13.8 - 30.0    100 Table 8: Log likelihood scores from STRUCTURE for Atlin Lake.  The boldface value represents the most likely K based on the maximum value of ∆K.  K log (K) ∆K 1 -4911.92 N/A 2 -4853.95 0.085968 3 -4792.66 0.387019 4 -4703.66 1.442054 5 -4655.1 2.957227 6 -4661.46 0.031059 7 -4670.67 0.036568 8 -4683.93 N/A   101  Table 9: Summary of allelic variation at eight microsatellite loci in lake trout of each subpopulation within Atlin Lake.  Number of alleles per locus (A), expected heterozygosity (He), observed heterozygosity (Ho), allelic richness(Ar), and number of individuals genotyped (N) for each loci per population.  Subpopulation SMM22 SCO102 SSA197 SCO19 SFO18 SCO107 SCO215 SCO2 Average A A 14 3 10 8 6 5 5 4 6.875 He 0.9056 0.8471 0.8471 0.7848 0.695 0.5795 0.5926 0.5227 0.6497 Ho 0.76 0.28 0.8182 0.7692 0.5714 0.4091 0.7391 0.4286 0.5970 Ar 13.465 2.84 9.907 7.422 6 4.999 4.907 4 6.6925 N 25 25 22 22 21 22 23 21 23.125 B A 19 3 12 7 6 7 4 5 7.875 He 0.9125 0.2549 0.8665 0.706 0.7707 0.7443 0.6465 0.4765 0.6746 Ho 1 0.26653 0.8511 0.5435 0.8293 0.6522 0.5652 0.4043 0.6419 Ar 14.641 2.96 10.281 5.756 5.971 5.828 3.918 4.415 6.7213 N 49 49 47 46 41 46 46 47 46.375 C A 19 3 13 10 6 9 7 6 9.125 He 0.9282 0.2284 0.8757 0.8091 0.7707 0.8109 0.5364 0.6021 0.7047 Ho 0.9677 0.1935 0.7742 0.7742 0.8293 0.7586 0.5357 0.3462 0.6241 Ar 16.95 2.677 11.789 9.204 6.938 8.171 6.323 5.581 8.4541 N 31 31 31 31 28 29 28 26 29.375 D A 17 3 13 10 7 9 3 4 8.25 He 0.8986 0.3523 0.823 0.8237 0.7908 0.7832 2540 0.5633 0.6446 Ho 0.8723 0.3696 0.7826 0.8511 0.6429 0.8889 0.2857 0.4444 0.6539 Ar 13.787 2.844 10.027 8.481 6.738 6.857 2.939 3.573 6.9058 N 47 46 46 47 38 45 42 36 43.375 E A 17 3 11 8 7 7 5 5 7.875 He 0.9011 0.3628 0.8506 0.8228 0.8221 0.7666 0.5708 0.5906 0.7173 Ho 0.9 0.3 0.9 0.7667 0.6429 0.75 0.5 0.5357 0.6619 Ar 15.391 3 9.989 7.687 6.941 6.744 4.691 4.736 7.3974 N 30 30 30 30 28 28 28 28 29 GLOBAL A 26 3 18 11 7 10 8 6 11.125 He 0.9092 0.4091 0.8526 0.7893 0.7699 0.7369 508.469 3 0.5510 64.1859 Ho 0.9000 0.2819 0.8252 0.7409 0.7032 0.6918 0.5251 0.4318 0.6375 Ar 17.6670 2.9060 11.3440 8.0090 6.7030 7.4650 5.0710 4.8900 8.0069 N 182 181 176 176 156 170 167 158 171.25    102  Table 10: Estimates of pairwise FST values (mean over 8 loci) for five lake trout subpopulations within Atlin Lake. Boldface values are significantly greater than 0.       A B C D B 0.074 C 0.049 0.038 D 0.067 0.058 0.034 E 0.059 0.032 0.011 0.032   103 Table 11: Estimates of effective population sizes of Atlin and Tagish lake subpopulations Subpopulation Sample Size  Ne 95% CI Atlin Lake A 26 29.1 20.6 - 45.8 B 50 334.5 126.0 - ∞ C 31 807.3 134.1 - ∞ D 47 101.8 65.1 - 208.7 E 30 116.9 60.7 - 727.6 Tagish Lake A 64 7641.9 339.4 - ∞ B 69 865.8 236.7 - ∞ C 35 94.3 59.8 - 200 D 59 524.9 187.5 - ∞      104 Table 12: Contingency test results of geographic units within Atlin Lake and Atlin Lake genetic subpopulations Geographic Units Genetic Subpopulations P (no association) = 0.0034873 Cramer’s V = 0.23061 Contigency C = 0.37093 81273E 129818D 94315C 2110118B 37313A 4321   105 Table 13: The mean pairwise identities for relatedness for Atlin Lake and Tagish Lake subpopulations (A-E and A-D, respectively). ATLIN Mean Identities  A B C D E A 0.312738 0.231994 0.224119 0.245352 0.218987 B  0.32216 0.240395 0.258389 0.252823 C   0.245398 0.244708 0.231000 D    0.326568 0.253901 E     0.26754  Standard Deviation  A B C D E A 0.167417 0.126314 0.115318 0.139895 0.123966 B  0.129873 0.115696 0.12107 0.123864 C   0.123715 0.117964 0.111293 D    0.138392 0.115907 E     0.13198  TAGISH Mean Identities  A B C D A 0.305228 0.307842 0.291897 0.275450 B  0.449446 0.309151 0.307158 C   0.309933 0.286959 D    0.319866  Standard Deviation  A B C D A 0.124661 0.123878 0.119445 0.115964 B  0.132493 0.124851 0.125176 C   0.127974 0.115900 D    0.12179     106 Table 14: Summary of allelic variation at eight microsatellite loci in lake trout of each subpopulation within Tagish Lake.  Number of alleles per locus (A), expected heterozygosity (He), observed heterozygosity (Ho), allelic richness(Ar) and number of individuals genotyped (N) for each loci per population.   Subpopulation SMM22 SCO102 SSA197 SCO19 SFO18 SCO107 SCO215 SCO2 Average A A 20 4 16 9 7 8 5 6 9.375 He 0.9089 0.4486 0.8846 0.7949 0.747 0.8192 0.4827 0.6184 0.7196 Ho 0.875 0.4762 0.9531 0.7302 0.5965 0.8095 0.4844 0.6774 0.8025 Ar 17.373 3.727 13.445 8.328 6.649 7.569 3.937 4.955 8.2479 N 64 63 64 63 57 63 64 62 62.5 B A 22 3 11 5 6 9 3 6 8.125 He 0.7955 0.1109 0.831 0.5305 0.5196 0.7965 0.437 0.6427 0.5928 Ho 0.8116 0.087 0.7826 0.4928 0.5538 0.75 0.4058 0.7612 0.5806 Ar 15.182 2.803 9.42 4.817 5.609 8.376 3 5.438 6.8306 N 69 69 69 69 65 68 69 67 68.125 C A 16 3 14 9 7 10 5 7 8.875 He 0.9016 0.209 0.8596 0.8188 0.7422 0.8441 0.6024 0.6322 0.7068 Ho 0.9714 0.2286 0.7714 0.8529 0.6667 0.8571 0.7143 0.8182 0.7174 Ar 15.492 3 13.124 8.763 7 9.571 4.963 6.72 8.5791 N 35 35 35 34 30 35 35 33 34 D A 19 3 13 8 7 9 6 5 8.75 He 0.9151 0.3503 0.7585 0.812 0.7624 0.7917 0.5331 0.5846 0.6995 Ho 0.8644 0.322 0.6724 0.6667 0.6667 0.7985 0.5345 0.661 0.6664 Ar 15.956 2.988 10.981 6.901 6.913 7.992 5.264 4.393 7.6735 N 59 59 58 51 59 59 58 59 57.75  GLOBAL A 23 4 19 10 7 11 6 10 11.25 He 0.8803 0.2797 0.8334 0.7391 0.6928 0.8129 0.5138 0.6195 0.6714 Ho 0.8806 0.2785 0.7949 0.6857 0.6209 0.8038 0.5348 0.7295 0.6661 Ar 17.1350 3.2220 12.5760 7.6100 6.5830 9.0320 4.4230 5.4880 8.2586 N 227 226 226 217 211 225 226 221 222.375   107  Table 15: Estimates of pairwise FST values (mean over 8 loci) for four lake trout subpopulations within Tagish Lake. Boldface values are significantly greater than 0.  A B C B 0.085 C 0.030 0.124 D 0.045 0.099 0.044     108  Table 16: Log likelihood scores from STRUCTURE for Tagish Lake. The boldface value represents the most likely K based on the maximum value of ∆K.  K log (K) ∆K 1 -6027.54 N/A 2 -5247.82 0.394178 3 -3440.93 0.858155 4 -5722.06 53.1892 5 -5723.36 0.343738 6 -5691.32 7.08578 7 -5091.98 0.45732 8 -5643.26 N/A         109 Table 17: Contingency test results of spawning locations within Tagish Lake and Tagish genetic subpopulations Spawning Locations Genetic Subpopulations P (no association) = 3.82 x10-15 Cramer’s V = 0.48896 Contingency C = 0.56876 22135TD 8121TC 61245TB 24172TA DBSWW   110 Table 18:  Log likelihood scores from STRUCTURE for Atlin and Tagish lakes. The boldface value represents the most likely K based on the maximum value of ∆K.  K log (K) ∆K 1 -11139.2 N/A 2 -11075.5 0.48235 3 -11044.6 1.723961 4 -11120.8 0.807367 5 -11298 0.012796 6 -11477.7 0.205563 7 -11725 0.570775 8 -11766.4 0.163916 9 -11898.2 5.512966 10 -14536.5 0.07708 11 -16392.7 0.09004 12 -17115.7 0.30011 13 -22095 N/A    111  Table 19: Loadings of eigenvectors on four principal components for the residuals of 36 body measurements on 104 lake trout from Atlin Lake  Character PC1 PC2 PC3 PC4 HD 0.031 -0.126 -0.461 -0.177 MBD 0.037 -0.097 -0.496 -0.135 CD 0.048 -0.053 -0.490 -0.168 Condition Factor 0.046 -0.053 -0.202 0.238 Brightness 0.036 0.061 0.168 -0.031 x1 0.044 0.304 -0.227 0.228 x2 0.153 -0.063 0.079 -0.295 x3 -0.059 0.254 -0.039 0.066 x4 -0.146 0.255 -0.112 0.214 x5 -0.201 0.020 0.040 0.076 x6 -0.088 -0.311 0.047 -0.099 x7 -0.077 -0.225 -0.139 0.068 x8 0.164 -0.075 -0.074 0.023 x9 0.188 0.143 0.051 0.069 x10 0.119 0.316 0.020 -0.084 x11 -0.085 0.317 0.041 -0.230 x12 -0.122 -0.116 -0.018 0.241 x13 -0.176 -0.014 -0.043 -0.071 x14 -0.099 -0.265 -0.006 0.140 x15 0.103 -0.196 0.264 -0.132 x16 0.190 -0.023 0.027 -0.111 y1 -0.246 0.036 0.006 0.049 y2 -0.161 0.157 -0.074 -0.059 y3 -0.234 -0.067 -0.010 0.136 y4 -0.230 -0.117 0.093 -0.027 y5 0.068 -0.311 0.056 -0.177 y6 0.241 -0.095 -0.040 0.042 y7 0.248 -0.064 -0.026 0.069 y8 0.211 0.008 -0.120 0.226 y9 -0.245 0.031 -0.017 -0.066 y10 -0.251 0.020 -0.007 -0.137 y11 -0.253 0.009 0.017 -0.085 y12 0.049 -0.081 0.009 0.505 y13 0.206 0.064 0.068 0.071 y14 0.239 0.060 0.060 0.014 y15 0.216 0.131 0.057 -0.143 y16 -0.050 0.237 -0.092 -0.229 % of variation  40.693 15.511 7.537 4.215     112 Table 20: Contingency test results assessing an association between geographic units and morphological groups in Atlin Lake.         16131802 39131521 4321 Geographic Units Morphological Clusters P (no association) = 0.0458 Cramer’s V = 0.2627 Contingency C = 0.2541   113 Table 21: Contingency test results assessing an association between genetic subpopulations and morphological groups in Atlin Lake.      11 741502 14 1982141 E DCBA Genetic  Subpopulations Morphological Clusters P (no association) = 0.3525 Cramer’s V = 0.2071 Contingency C = 0.2028   114 Table 22: Genetic mixture analysis results of 101 commercial samples and 33 recreational angling samples of lake trout within Atlin Lake. Mixture values represent the estimates contribution of each subpopulation within Atlin and Tagish lakes. The averages and standard deviations are based on simulations from 5000 replicates. The boldface values represent the subpopulations with the greatest estimated contributions.                  Commercial                          Angling  Mixture Avg P St Dev Mixture Avg P St Dev Atlin - A 0.0330 0.0358 0.0315 0.1203 0.1293 0.0541 Atlin - B 0.2146 0.2127 0.0640 0.1002 0.0998  0.0396 Atlin - C 0.1763 0.1729 0.0559 0.2508 0.2529  0.0527 Atlin - D 0.0458 0.0499 0.0411 0.1352 0.1202  0.0441 Atlin - E 0.1331 0.1358 0.0588 0.1409 0.1269  0.0435 TAG - A 0.1643 0.1608 0.0723 0.0523 0.0610  0.0563 TAG - B 0.0358 0.0369 0.0312 0.0000 0.0090  0.0034 TAG - C 0.1280 0.1247 0.0588 0.0000 0.0033  0.0093 TAG - D 0.0690 0.0704 0.0471 0.2004  0.1976  0.0721   115 Table 23: Genetic mixture analysis results of 101 commercial samples and 33 recreational angling samples of lake trout within Atlin Lake. Mixture values represent the estimates contribution of each subpopulation within Atlin and Tagish lakes. The averages and standard deviations are based on simulations from 5000 replicates.The boldface values represent the subpopulations with the greatest estimated contributions. The mixture analysis was conducted using rigid reference populations with admixed genomes removed (> 0.5 admixed) for both Atlin and Tagish subpopulations.                           Commercial                          Angling  Mixture Avg P St Dev Mixture Avg P St Dev Atlin - A 0.0307 0.0309 0.0254 0.1341 0.1339 0.0496 Atlin - B 0.2252 0.2233 0.0617 0.0488 0.0515 0.0343 Atlin - C 0.1928 0.1905 0.0584 0.2758 0.2736 0.0665 Atlin - D 0.0679 0.0683 0.0398 0.0956 0.0928 0.0443 Atlin - E 0.0607 0.0622 0.0365 0.0841 0.0862 0.0427 TAG - A 0.1705 0.1706 0.0586 0.0729 0.0755 0.0423 TAG - B 0.0312 0.0310 0.0259 0.0000 0.0011 0.0045 TAG - C 0.1280 0.1305 0.0521 0.0000 0.0048 0.0109 TAG - D 0.0930 0.0926 0.0463 0.2879 0.2807 0.0669       116 TV F M AL TAG T TR KL TE P CL N H B A K AR FRA Pacific Ocean CU  Figure 1: Location of lake trout sample sites examined in this study.  Population codes are given in Table 1.   117  Smm22 0 0.1 0.2 0.3 0.4 0.5 13 2 13 6 14 0 14 4 14 8 15 2 16 0 16 4 16 8 17 2 17 6 18 0 18 4 18 8 19 0 19 2 19 4 19 6 19 8 20 2 20 6 21 0 21 4 21 8 22 2 22 6 23 0 23 4 23 6 23 8 Allele Size (base pairs) A lle le  F re qu en cy Yukon Taku Mackenzie Fraser Sco102 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 144 168 172 176 Allele Size (base pairs) A lle le  F re qu en cy Yukon Taku Mackenzie Fraser A B   118  D Ssa197 0 0.1 0.2 0.3 199 219 223 227 231 235 239 243 247 251 255 259 263 267 271 275 279 283 287 291 295 299 303 307 Allele Size (base pairs) A lle le  F re qu en cy Yukon Taku Mackenzie Fraser Sco19 0 0.1 0.2 0.3 0.4 0.5 144 154 156 158 160 164 166 168 170 172 174 176 178 180 182 Allele Size (base pairs) A lle le  F re qu en cy Yukon Taku Mackenzie Fraser C   119   E F Sfo18 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 162 166 170 174 178 182 186 Allele Size (base pairs) A lle le  F re qu en cy Yukon Taku Mackenzie Fraser Sco107 0 0.1 0.2 0.3 0.4 0.5 200 204 208 212 216 220 224 228 232 236 240 244 248 252 256 272 Allele Size (base pairs) A lle le  F re qu en cy Yukon Taku Mackenzie Fraser   120  Figure 2: Allele frequencies in lake trout from four major watersheds and assayed at eight microsatellite loci G H Sco215 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 283 291 295 299 303 307 311 315 319 Allele Size (base pairs) A lle le  F re qu en cy Yukon Taku Mackenzie Fraser Sco2 0 0.1 0.2 0.3 0.4 0.5 0.6 161 167 169 171 173 177 179 181 183 185 187 189 207 Allele Size (base pairs) A lle le  F re qu en cy Yukon Taku Mackenzie Fraser   121       12 .9 9%  o f v ar ia tio n,  P  =  0 .2 90 7 17 18 118 2 5 6 41 3 12 14 16 9 151910 13 51.58% of variation, P = 0.010 12 .9 9%  o f v ar ia tio n,  P  =  0 .2 90   Figure 3: Principal components analysis of allele frequency variation at eight microsatellite loci among all sample lakes.  The numbers correspond to the population codes listed in Table 1.           Fraser River Watershed Yukon River Watershed   122  A. All Populations (N=19) R = 0.2694 p=0.0910 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 500 1000 1500 2000 G e n e t i c  D i s t a n c e  ( F s t ) B. Fraser Watershed (N=12) R = 0.4710 p=0.0250 0 50 100 150 200 250 300 350 0 0.1 0.2 0.3 0.4 0.5 0.6 Geographic Distance (km) G e n e t i c  D i s t a n c e  ( F s t ) C. PCAGroup1 (N=6) R = 0.3370 p=0.2450 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0 100 200 300 400 500 600 D. PCAGroup2 (N=13) R =0.2441 p=0.2200 0 0.1 0.2 0.3 0.4 0.5 0.6 0 500 1000 1500 2000 Geographic Distance (km) G e n e t i c  D i s t a n c e  ( F s t ) G e n e t i c  D i s t a n c e  ( F s t )  Figure 4: Isolation-by-distance analyses for lake trout populations in western Canada. r = 0.27 P = 0.09 r = 0.34 P = 0.25 r = 0.47 P = 0.03 r  0.24 P = 0.22   123  Smm22 0 0.05 0.1 0.15 0.2 0.25 13 6 14 0 14 4 14 8 15 2 16 0 16 4 16 8 17 2 17 6 18 0 18 4 18 8 19 0 19 2 19 4 19 6 19 8 20 2 20 6 21 0 21 4 21 8 22 2 22 6 23 4 Allele Sizes (Base Pairs) A lle le  F re qu en cy A B C D E Sco102 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 144 172 176 Allele Sizes (Base Pairs) A lle le  F re qu en cy A B C D E A B   124   Ssa197 0 0.05 0.1 0.15 0.2 0.25 0.3 199 219 223 227 231 235 239 243 247 251 255 259 263 267 271 275 279 287 Allele Sizes (Base Pairs) A lle le  F re qu en cy A B C D E Sco19 0 0.1 0.2 0.3 0.4 0.5 144 154 156 158 160 164 166 168 170 172 176 Allele Sizes (Base Pairs) A lle le  F re qu en cy A B C D E D C   125  Sfo18 0 0.1 0.2 0.3 0.4 0.5 162 166 170 174 178 182 186 Allele Sizes (Base Pairs) A lle le  F re qu en cy A B C D E Sco107 0 0.1 0.2 0.3 0.4 0.5 0.6 200 204 208 212 216 232 236 240 244 256 Allele Sizes (Base Pairs) A lle le  F re qu en cy A B C D E F E   126  Figure 5: Allele frequencies in lake trout sampled from Atlin Lake and assayed at eight microsatellite loci.  A-E refer to the five subpopulations found in Atlin Lake and correspond to those mentioned in Table 9. H Sco215 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 283 291 299 303 307 311 315 319 Allele Sizes (Base Pairs) A lle le  F re qu en cy  A B C D E Sco2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 171 173 179 181 183 185 Allele Sizes (Base Pairs) A lle le  F re qu en cy  A B C D E G   127 A B C D E 57 42 31 53   Figure 6: Map of the distribution of lake trout subpopulations within Atlin Lake according to STRUCTURE assignments.  The ----- indicates the separation between geographical units.  The value to the right is the sample size for each section of the lake.   128 B Smm22 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 136 140 160 164 172 176 180 184 188 190 192 194 196 198 202 206 210 214 218 222 226 230 238 Allele Size (Base Pairs) A lle le  F re qu en cy TA TB TC TD Sco102 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 144 168 172 176 Allele Size (Base Pairs) A lle le  F re qu en cy  TA TB TC TD A   129  D Ssa197 0 0.1 0.2 0.3 199 219 223 227 231 235 239 243 247 251 255 259 263 267 271 275 279 283 287 Allele Size (Base Pairs) A lle le  F re qu en cy TA TB TC TD Sco19 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 144 154 156 158 160 164 166 168 170 172 Allele Size (Base Pairs) A lle le  F re qu en cy TA TB TC TD C   130 F E Sfo18 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 162 166 170 174 178 182 186 Allele Size (Base Pairs) A lle le  F re qu en cy TA TB TC TD Sco107 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 200 204 208 212 216 232 236 240 244 248 252 Allele Size (Base Pairs) A lle le  F re qu en cy TA TB TC TD   131  Figure 7: Allele frequencies in lake trout sampled from Tagish Lake and assayed at eight microsatellite loci. TA-TD refer to the four subpopulations found in Tagish Lake and correspond those mentioned in Table 13. H G Sco215 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 283 295 299 303 311 315 Allele Size (Base Pairs) A lle le  F re qu en cy TA TB TC TD Sco2 0 0.1 0.2 0.3 0.4 0.5 0.6 167 169 171 173 177 179 181 183 185 189 Allele Size (Base Pairs) A lle le  F re qu en cy TA TB TC TD   132    6 1 2 4 29.95% of variation, P = 0.060 20 .6 1%  o f v ar ia tio n,  P  =  0 .1 30 20 .6 1%  o f v ar ia tio n,  P  =  0 .1 30 7 59 2 3 8 1 4 6 20 .6 1%  o f v ar ia tio n,  P  =  0 .1 30 20 .6 1%  o f v ar ia tio n,  P  =  0 .1 30   Figure 8: Principal components analysis of allele frequency variation at eight microsatellite loci among subpopulations of lake trout in Atlin and Tagish Lake (1-5 corresponds with Atlin subpopulations A-E and 6-9 with Tagish subpopulations A-D).   133  Tagish Lake SW (56) DB (62) W (55) TA TB TC TD 4 0 3   Figure 9: Map of the distribution of lake trout subpopulations at spawning beds within Tagish Lake according to STRUCTURE assignments.  The value to the right is the sample size for each section of the lake.     134  Figure 10: Known or Suspected Spawning Locations in Atlin Lake and Tagish Lake (Atlin Community Fisheries Working Group 2001) Tagish Lake Atlin Lake Atlin River SW DB SC W   135    Figure 11:  Locations of colouration measurements used in morphological analysis.  Each square represents an area 40x40psi.   136   Figure 12: Landmarks used to compare shapes of lake trout sampled within Atlin Lake: 1 -  anterior tip of snout, 2 - posterior tip of maxilla, 3 – center of eye, 4 – top of cranium at midpoint of eye, 5 – posterior of neurocranium above tip of opercle, 6 – anterior insertion of dorsal fin, 7 – posterior insertion of dorsal fin, 8 – anterior insertion of adipose fin, 9 – dorsal insertion of caudal fin, 10 – midpoint of hypural plate, 11 – ventral insertion of caudal fin, 12 – posterior insertion of anal fin, 13 – anterior insertion of anal fin, 14 – insertion point of pelvic fin, 15 – insertion point of pectoral fin, 16 – connection between gill covers.     137   Figure 13: Head depth (HD), mid body depth (MBD), caudal peduncle depth (CD), and body length (FL) measurements of lake trout from Atlin Lake.     138 1 3 2 4 6 5 7 8 9 10 1116 15 13 12 14   Figure 14:  Transitions in landmark position contributing to PC1 of morphological variation in lake trout.  Each arrow is an exaggeration of the displacement in the X and Y coordinates of each landmark position.        139  Figure 15: Models generated by McCLUST to determine the most probable number of clusters (components) within the lake trout morphological data based on Bayesian Information Criterion (BIC).   Each line and symbol represents a different model (see McCLUST for model definitions).   140    Figure 16:  Atlin Lake lake trout samples plotted along the first and second principle components (PC1 and P2).  Ellipses encircle about 95% of the measurements present in each cluster of individuals as classified in the MCLUST analysis.  Triangles correspond to morphotype 1 and squares to morphotype 2.   141   Figure 17: Map of the distribution of the two morphotypes among four geographic localities  within Atlin Lake.   The value to the right is the sample size for each section of the lake.    142   Figure 18:  Results of DFA indicating the level of overlap between the two morphotypes. Morphotype 1 Morphotype 2 -4 -3.2 -2.4 -1.6 -0.8 0 0.8 1.6 2.4 3.2 Discriminant 0 3 6 9 12 15 18 21 24 27 F re qu en cy   143 39112.252 69121.781 NMeanMorphology Cluster 28126.21E 32114.58D 16100.49C 38121.76B 7115.83A NMeanGenetic Subpopulation 57127.424 33109.583 31108.102 7108.661 NMeanGeographic Unit a) Colouration in genetic subpopulations b) Colouration in morphological clusters c) Colouration in geographic units 1 2 3 4 0   Black 256  White ANOVA F = 1.816  P = 0.130 ANOVA F = 2.134 P = 0.147 ANOVA F = 3.568  P = 0.016 Figure 19:  ANOVA results of colouration comparisons to (a) genetic subpopulations (b) morphological clusters and (c) geographic unit.  Map indicates the geographic units within Atlin Lake.   Pictures indicate the extreme colouration differences found within Atlin Lake.   144  Figure 20: Map of the commercial fishery (squares) and common angling fishing (circles) locations within Atlin Lake  Great Northern Fish Company Atlin River

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0066241/manifest

Comment

Related Items